In order to avoid security-related warning messages when switching to secured connection, you may want either to:

Click here to proceed.

This website allows you to compute physicochemical descriptors as well as to predict ADME parameters, pharmacokinetic properties, druglike nature and medicinal chemistry friendliness of one or multiple small molecules to support drug discovery.

The main article describing the web service and its underlying methodologies is SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. (2017) 7:42717.
For details about development and validation of iLOG, please refer to this article: iLOGP: a simple, robust, and efficient description of n-octanol/water partition coefficient for drug design using the GB/SA approach. J. Chem. Inf. Model. (2014) 54(12):3284-3301.
For details about development and validation of the BOILED-Egg, please refer to this article: A BOILED-Egg to predict gastrointestinal absorption and brain penetration of small molecules. ChemMedChem (2016) 11(11):1117-1121.

Developed and maintained by the Molecular Modeling Group of the SIB | Swiss Institute of Bioinformatics.


Enter a list of SMILES here:

Retrieve data:    


0
20
40
60
80
100
120
140
160
180
TPSA
0
1
2
3
4
5
6
7
-4
-3
-2
-1
WLOGP
Actions
Show Molecules Name

Legends
BBB
HIA
PGP+
PGP—
Remarks
20 molecules out of range!
Molecule 1
SMILES [O-]C(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)ccc2=O
Physicochemical Properties
Formula C14H5O6
Molecular weight 269.19 g/mol
Num. heavy atoms 20
Num. arom. heavy atoms 13
Fraction Csp3 0.00
Num. rotatable bonds 1
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 66.59
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

111.57 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.84
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.78
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.46
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.42
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.48
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.13
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.43
Solubility 9.94e+00 mg/ml ; 3.69e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.08
Solubility 2.21e+01 mg/ml ; 8.23e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-3.84
Solubility 3.90e-02 mg/ml ; 1.45e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.50 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.16
Molecule 2
SMILES SC(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C16H7O7S
Molecular weight 343.29 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 3
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 84.49
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

167.44 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.05
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.80
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.46
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.78
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.08
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.22
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.67
Solubility 7.39e+00 mg/ml ; 2.15e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.24
Solubility 1.99e+00 mg/ml ; 5.79e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.27
Solubility 1.86e-02 mg/ml ; 5.41e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.96 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

2 alerts: thioester, thiol_2

Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.33
Molecule 3
SMILES NC(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C16H8NO7
Molecular weight 326.24 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 3
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 79.27
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

154.66 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.04
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.96
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.43
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.17
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.08
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.69
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.83
Solubility 4.82e+01 mg/ml ; 1.48e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.77
Solubility 5.60e+01 mg/ml ; 1.72e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-3.81
Solubility 5.08e-02 mg/ml ; 1.56e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.68 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.29
Molecule 4
SMILES [N+]CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C16H6NO7
Molecular weight 324.22 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 3
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 77.86
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

128.64 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-4.07
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.21
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.26
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.08
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.89
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

0.51
Solubility 1.05e+03 mg/ml ; 3.24e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

1.97
Solubility 3.03e+04 mg/ml ; 9.35e+01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-3.81
Solubility 5.04e-02 mg/ml ; 1.56e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.17 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.24
Molecule 5
SMILES CCCC[N+]CC(=O)C1=CC(=O)C2=C(C1=O)C1=C(O)C=CC(=O)C1C2C(=O)[O-]
Physicochemical Properties
Formula C20H17NO7
Molecular weight 383.35 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 0
Fraction Csp3 0.35
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 94.35
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

128.64 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-11.04
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.72
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.60
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.61
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-3.45
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.16
Solubility 2.65e+02 mg/ml ; 6.90e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.37
Solubility 9.04e+02 mg/ml ; 2.36e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-2.65
Solubility 8.48e-01 mg/ml ; 2.21e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.43 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: quinone_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

2 alerts: chinone_1, michael_acceptor_4

Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.71
Molecule 6
SMILES OCC[N+]CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C18H11NO8
Molecular weight 369.28 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 13
Fraction Csp3 0.17
Num. rotatable bonds 6
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 89.69
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

148.87 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-12.98
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-4.25
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.68
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.55
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.34
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-4.62
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

0.59
Solubility 1.43e+03 mg/ml ; 3.87e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

1.73
Solubility 2.00e+04 mg/ml ; 5.40e+01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.42
Solubility 1.39e-02 mg/ml ; 3.77e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.57 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.48
Molecule 7
SMILES C[N+](CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C
Physicochemical Properties
Formula C18H12NO7
Molecular weight 354.29 g/mol
Num. heavy atoms 26
Num. arom. heavy atoms 13
Fraction Csp3 0.17
Num. rotatable bonds 4
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 89.59
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-4.58
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-3.10
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.88
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.78
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.48
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-2.37
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.19
Solubility 2.29e+02 mg/ml ; 6.46e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.90
Solubility 2.79e+03 mg/ml ; 7.87e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.26
Solubility 1.95e-02 mg/ml ; 5.50e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.66 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.48
Molecule 8
SMILES CC[N+](CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CC
Physicochemical Properties
Formula C20H16NO7
Molecular weight 382.34 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.25
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 99.20
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.96
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.36
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.10
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.32
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.29
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.69
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.67
Solubility 8.15e+01 mg/ml ; 2.13e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.13
Solubility 5.14e+02 mg/ml ; 1.34e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.05
Solubility 3.42e-03 mg/ml ; 8.95e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.31 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.70
Molecule 9
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCCCC1
Physicochemical Properties
Formula C21H16NO7
Molecular weight 394.35 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.29
Num. rotatable bonds 4
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 105.51
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-4.23
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.25
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.09
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.38
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.70
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.94
Solubility 4.58e+01 mg/ml ; 1.16e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.01
Solubility 4.07e+02 mg/ml ; 1.03e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.84
Solubility 5.67e-03 mg/ml ; 1.44e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.30 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.71
Molecule 10
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCOCC1
Physicochemical Properties
Formula C20H14NO8
Molecular weight 396.33 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.25
Num. rotatable bonds 4
Num. H-bond acceptors 9
Num. H-bond donors 1
Molar Refractivity 101.79
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

141.11 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-4.32
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-3.47
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.49
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.09
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.75
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-2.52
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.18
Solubility 2.63e+02 mg/ml ; 6.62e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

1.09
Solubility 4.83e+03 mg/ml ; 1.22e+01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.30
Solubility 1.98e-02 mg/ml ; 4.99e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.18 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.65
Molecule 11
SMILES O=C1C(=C(C1=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])O
Physicochemical Properties
Formula C18H5O9
Molecular weight 365.23 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 13
Fraction Csp3 0.00
Num. rotatable bonds 2
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 86.15
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

165.94 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-0.14
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.63
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.01
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.58
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.65
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.30
Solubility 1.82e+01 mg/ml ; 4.99e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.34
Solubility 1.65e+01 mg/ml ; 4.52e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.07
Solubility 3.11e-02 mg/ml ; 8.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.69 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.50
Molecule 12
SMILES O=C1C(=C(C1=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])O
Physicochemical Properties
Formula C19H7O9
Molecular weight 379.25 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.05
Num. rotatable bonds 3
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 90.16
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

165.94 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.33
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.33
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.95
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.78
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.97
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.35
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.50
Solubility 1.20e+01 mg/ml ; 3.17e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.66
Solubility 8.38e+00 mg/ml ; 2.21e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.46
Solubility 1.30e-02 mg/ml ; 3.44e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.56 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.54
Molecule 13
SMILES O=C1NC(=O)C(S1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C17H6NO8S
Molecular weight 384.30 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 2
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 94.15
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

183.04 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.48
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.92
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.14
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.96
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.31
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.25
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.87
Solubility 5.22e+00 mg/ml ; 1.36e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.44
Solubility 1.39e+00 mg/ml ; 3.63e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.31
Solubility 1.87e-02 mg/ml ; 4.88e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.30 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.86
Molecule 14
SMILES O=C1NC(=O)C(S1)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C18H8NO8S
Molecular weight 398.32 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.11
Num. rotatable bonds 3
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 98.96
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

183.04 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.39
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.63
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.95
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.72
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.04
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.06
Solubility 3.48e+00 mg/ml ; 8.74e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.74
Solubility 7.23e-01 mg/ml ; 1.81e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.71
Solubility 7.86e-03 mg/ml ; 1.97e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.18 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.92
Molecule 15
SMILES Oc1ccc(=O)c2c1c1=c(c2C(=O)[O-])c(=O)cc(c1=O)c1ncsc1
Physicochemical Properties
Formula C17H6NO6S
Molecular weight 352.30 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 18
Fraction Csp3 0.00
Num. rotatable bonds 2
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 87.69
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

152.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.61
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.08
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.64
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.15
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.74
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.37
Solubility 1.49e+00 mg/ml ; 4.22e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.67
Solubility 7.45e-01 mg/ml ; 2.11e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.21
Solubility 2.17e-03 mg/ml ; 6.15e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.51 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.31
Molecule 16
SMILES Oc1ccc(=O)c2c1c1=c(c2C(=O)[O-])c(=O)cc(c1=O)Cc1cscn1
Physicochemical Properties
Formula C18H8NO6S
Molecular weight 366.32 g/mol
Num. heavy atoms 26
Num. arom. heavy atoms 18
Fraction Csp3 0.06
Num. rotatable bonds 3
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 91.71
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

152.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.06
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.22
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.41
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.41
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.53
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.00
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.56
Solubility 9.99e-01 mg/ml ; 2.73e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.99
Solubility 3.78e-01 mg/ml ; 1.03e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.61
Solubility 9.10e-04 mg/ml ; 2.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.38 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.40
Molecule 17
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccccc1
Physicochemical Properties
Formula C22H12NO7
Molecular weight 402.33 g/mol
Num. heavy atoms 30
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 105.61
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.11
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.83
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.35
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.62
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.02
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.00
Solubility 4.06e-01 mg/ml ; 1.01e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.37
Solubility 1.73e-01 mg/ml ; 4.30e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.68
Solubility 8.41e-05 mg/ml ; 2.09e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.52
Molecule 18
SMILES Cc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C23H14NO7
Molecular weight 416.36 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 110.57
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.76
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.19
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.05
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.09
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.15
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.23
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.29
Solubility 2.11e-01 mg/ml ; 5.07e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.74
Solubility 7.57e-02 mg/ml ; 1.82e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.05
Solubility 3.68e-05 mg/ml ; 8.83e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.99 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.63
Molecule 19
SMILES Clc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C22H11ClNO7
Molecular weight 436.78 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 110.62
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.09
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.45
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.30
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.36
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.27
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.49
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.59
Solubility 1.14e-01 mg/ml ; 2.60e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.01
Solubility 4.27e-02 mg/ml ; 9.77e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.26
Solubility 2.38e-05 mg/ml ; 5.45e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.93 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.51
Molecule 20
SMILES Brc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C22H11BrNO7
Molecular weight 481.23 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 113.31
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.08
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.52
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.41
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.46
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.31
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.56
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.90
Solubility 5.99e-02 mg/ml ; 1.25e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.08
Solubility 3.98e-02 mg/ml ; 8.27e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.45
Solubility 1.69e-05 mg/ml ; 3.52e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MW>480
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.54
Molecule 21
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C23H11Cl3NO7
Molecular weight 519.69 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 124.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.78
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.45
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.36
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.98
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.39
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.62
Solubility 1.24e-02 mg/ml ; 2.38e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.05
Solubility 4.66e-03 mg/ml ; 8.96e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.46
Solubility 1.82e-06 mg/ml ; 3.51e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.73 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MW>480
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.64
Molecule 22
SMILES Oc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C22H12NO8
Molecular weight 418.33 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 107.63
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

160.90 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.33
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.47
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.65
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.64
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.15
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.53
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.85
Solubility 5.86e-01 mg/ml ; 1.40e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.42
Solubility 1.60e-01 mg/ml ; 3.82e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.09
Solubility 3.40e-04 mg/ml ; 8.13e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.52 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.56
Molecule 23
SMILES COc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C23H14NO8
Molecular weight 432.36 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 112.10
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

149.90 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.74
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.80
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.35
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.42
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.89
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.07
Solubility 3.70e-01 mg/ml ; 8.55e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.53
Solubility 1.28e-01 mg/ml ; 2.96e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.78
Solubility 7.18e-05 mg/ml ; 1.66e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.37 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.64
Molecule 24
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C22H12N2O9
Molecular weight 448.34 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 6
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 110.68
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.98 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.19
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.55
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.14
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.20
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.77
Solubility 7.62e-01 mg/ml ; 1.70e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.55
Solubility 1.27e-01 mg/ml ; 2.83e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.02
Solubility 4.31e-04 mg/ml ; 9.62e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.90 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.63
Molecule 25
SMILES [O-]C(=O)c1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C23H11NO9
Molecular weight 445.33 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 110.62
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.80 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.27
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.35
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.99
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.55
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.11
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.04
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.85
Solubility 6.27e-01 mg/ml ; 1.41e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.71
Solubility 8.66e-02 mg/ml ; 1.94e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.01
Solubility 4.33e-04 mg/ml ; 9.73e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.77 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.60
Molecule 26
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H16NO9
Molecular weight 474.40 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 19
Fraction Csp3 0.12
Num. rotatable bonds 8
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 121.69
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

166.97 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.09
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.05
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.18
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.03
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.07
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.20
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.32
Solubility 2.29e-01 mg/ml ; 4.83e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.15
Solubility 3.38e-02 mg/ml ; 7.13e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.09
Solubility 3.85e-05 mg/ml ; 8.11e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.45 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.85
Molecule 27
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1cccc2c1cccc2
Physicochemical Properties
Formula C26H14NO7
Molecular weight 452.39 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 23
Fraction Csp3 0.04
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 123.11
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.75
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.08
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.80
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.55
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.66
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.77
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.13
Solubility 3.39e-02 mg/ml ; 7.48e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.66
Solubility 9.81e-03 mg/ml ; 2.17e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.31
Solubility 2.23e-06 mg/ml ; 4.92e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.58 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.74
Molecule 28
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C26H14NO7
Molecular weight 452.39 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 23
Fraction Csp3 0.04
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 123.11
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.21
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.08
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.80
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.55
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.66
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.86
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.13
Solubility 3.39e-02 mg/ml ; 7.48e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.66
Solubility 9.81e-03 mg/ml ; 2.17e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.31
Solubility 2.23e-06 mg/ml ; 4.92e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.58 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.74
Molecule 29
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccccc1
Physicochemical Properties
Formula C25H14N3O7S
Molecular weight 500.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.94
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.36
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.70
Solubility 1.00e-01 mg/ml ; 2.00e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.81
Solubility 7.66e-03 mg/ml ; 1.53e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.50e-06 mg/ml ; 8.99e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.54 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.92
Molecule 30
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)C
Physicochemical Properties
Formula C26H16N3O7S
Molecular weight 514.49 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.80
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.75
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.64
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.08
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.02
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.80
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.00
Solubility 5.16e-02 mg/ml ; 1.00e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.19
Solubility 3.33e-03 mg/ml ; 6.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.42
Solubility 1.97e-06 mg/ml ; 3.83e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.37 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.03
Molecule 31
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)Cl
Physicochemical Properties
Formula C25H13ClN3O7S
Molecular weight 534.90 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.60
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.77
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.99
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.08
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.14
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.91
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.29
Solubility 2.75e-02 mg/ml ; 5.13e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.46
Solubility 1.86e-03 mg/ml ; 3.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.62
Solubility 1.28e-06 mg/ml ; 2.39e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.31 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.90
Molecule 32
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)Br
Physicochemical Properties
Formula C25H13BrN3O7S
Molecular weight 579.36 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 138.53
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.84
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.84
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.10
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.18
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.17
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.03
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.61
Solubility 1.42e-02 mg/ml ; 2.46e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.53
Solubility 1.71e-03 mg/ml ; 2.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.80
Solubility 9.14e-07 mg/ml ; 1.58e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.53 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.92
Molecule 33
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C26H13Cl3N3O7S
Molecular weight 617.82 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 150.07
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.24
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.77
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.05
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.67
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.26
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.80
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.33
Solubility 2.88e-03 mg/ml ; 4.66e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.50
Solubility 1.97e-04 mg/ml ; 3.19e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.79
Solubility 9.91e-08 mg/ml ; 1.60e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.10 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: MW>600, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.04
Molecule 34
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)O
Physicochemical Properties
Formula C25H14N3O8S
Molecular weight 516.46 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 4
Molar Refractivity 132.86
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

214.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.30
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.79
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.62
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.02
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.11
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.43e-01 mg/ml ; 2.77e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.87
Solubility 7.03e-03 mg/ml ; 1.36e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.45
Solubility 1.82e-05 mg/ml ; 3.52e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.89 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 35
SMILES COc1ccc(cc1)NCC(=O)Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H16N3O8S
Molecular weight 530.49 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 137.33
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

203.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.94
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.34
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.42
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.57
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.51
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.77
Solubility 8.93e-02 mg/ml ; 1.68e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.98
Solubility 5.58e-03 mg/ml ; 1.05e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.14
Solubility 3.85e-06 mg/ml ; 7.26e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.74 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.05
Molecule 36
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C25H14N4O9S
Molecular weight 546.47 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 8
Num. H-bond acceptors 10
Num. H-bond donors 4
Molar Refractivity 135.91
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

234.14 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.13
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.36
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.36
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.53
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.48
Solubility 1.82e-01 mg/ml ; 3.34e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.00
Solubility 5.50e-03 mg/ml ; 1.01e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.37
Solubility 2.32e-05 mg/ml ; 4.24e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.27 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.03
Molecule 37
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C26H13N3O9S
Molecular weight 543.46 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 8
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 135.85
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

233.96 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.81
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.67
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.30
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.80
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.98
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.47
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.50e-01 mg/ml ; 2.76e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.16
Solubility 3.76e-03 mg/ml ; 6.92e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.37
Solubility 2.33e-05 mg/ml ; 4.29e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.14 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 38
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C28H18N3O9S
Molecular weight 572.52 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 10
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 146.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

220.13 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.36
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.51
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.29
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.94
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.52
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.02
Solubility 5.47e-02 mg/ml ; 9.56e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.59
Solubility 1.49e-03 mg/ml ; 2.60e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.44
Solubility 2.08e-06 mg/ml ; 3.63e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.83 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.25
Molecule 39
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1cccc2c1cccc2
Physicochemical Properties
Formula C29H16N3O7S
Molecular weight 550.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 28
Fraction Csp3 0.03
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 148.34
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.49
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.51
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.50
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.29
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.82
Solubility 8.31e-03 mg/ml ; 1.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.11
Solubility 4.25e-04 mg/ml ; 7.72e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.66
Solubility 1.20e-07 mg/ml ; 2.18e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.14
Molecule 40
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C29H16N3O7S
Molecular weight 550.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 28
Fraction Csp3 0.03
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 148.34
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.09
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.49
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.51
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.50
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.40
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.82
Solubility 8.31e-03 mg/ml ; 1.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.11
Solubility 4.25e-04 mg/ml ; 7.72e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.66
Solubility 1.20e-07 mg/ml ; 2.18e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.14
Molecule 41
SMILES O=C(Nc1ccccc1)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H14N3O7S
Molecular weight 500.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.93
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.55
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.70
Solubility 1.00e-01 mg/ml ; 2.00e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.81
Solubility 7.66e-03 mg/ml ; 1.53e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.50e-06 mg/ml ; 8.99e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.54 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 42
SMILES O=C(Nc1ccc(cc1)C)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H16N3O7S
Molecular weight 514.49 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.80
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.16
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.64
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.08
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.02
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.88
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.00
Solubility 5.16e-02 mg/ml ; 1.00e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.19
Solubility 3.33e-03 mg/ml ; 6.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.42
Solubility 1.97e-06 mg/ml ; 3.83e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.37 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 43
SMILES O=C(Nc1ccc(cc1)Cl)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H13ClN3O7S
Molecular weight 534.90 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.51
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.77
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.99
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.08
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.14
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.90
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.29
Solubility 2.75e-02 mg/ml ; 5.13e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.46
Solubility 1.86e-03 mg/ml ; 3.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.62
Solubility 1.28e-06 mg/ml ; 2.39e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.31 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.93
Molecule 44
SMILES O=C(Nc1ccc(cc1)Br)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H13BrN3O7S
Molecular weight 579.36 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 138.53
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.05
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.84
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.10
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.18
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.17
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.07
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.61
Solubility 1.42e-02 mg/ml ; 2.46e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.53
Solubility 1.71e-03 mg/ml ; 2.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.80
Solubility 9.14e-07 mg/ml ; 1.58e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.53 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 45
SMILES O=C(Nc1ccc(cc1)C(Cl)(Cl)Cl)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H13Cl3N3O7S
Molecular weight 617.82 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 150.07
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.77
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.05
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.67
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.26
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.77
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.33
Solubility 2.88e-03 mg/ml ; 4.66e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.50
Solubility 1.97e-04 mg/ml ; 3.19e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.79
Solubility 9.91e-08 mg/ml ; 1.60e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.10 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: MW>600, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.07
Molecule 46
SMILES O=C(Nc1ccc(cc1)O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H14N3O8S
Molecular weight 516.46 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 4
Molar Refractivity 132.86
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

214.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.83
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.79
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.89
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.02
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.16
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.43e-01 mg/ml ; 2.77e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.87
Solubility 7.03e-03 mg/ml ; 1.36e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.45
Solubility 1.82e-05 mg/ml ; 3.52e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.89 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 47
SMILES COc1ccc(cc1)NC(=O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H16N3O8S
Molecular weight 530.49 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 137.33
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

203.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.77
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.34
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.68
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.57
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.42
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.77
Solubility 8.93e-02 mg/ml ; 1.68e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.98
Solubility 5.58e-03 mg/ml ; 1.05e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.14
Solubility 3.85e-06 mg/ml ; 7.26e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.74 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.08
Molecule 48
SMILES O=C(Nc1ccc(cc1)[N+](=O)O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C25H14N4O9S
Molecular weight 546.47 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 8
Num. H-bond acceptors 10
Num. H-bond donors 4
Molar Refractivity 135.91
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

234.14 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.13
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.36
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.36
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.53
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.48
Solubility 1.82e-01 mg/ml ; 3.34e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.00
Solubility 5.50e-03 mg/ml ; 1.01e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.37
Solubility 2.32e-05 mg/ml ; 4.24e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.27 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 49
SMILES O=C(Nc1ccc(cc1)C(=O)[O-])CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H13N3O9S
Molecular weight 543.46 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 8
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 135.85
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

233.96 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.90
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.67
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.30
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.69
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.98
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.71
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.50e-01 mg/ml ; 2.76e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.16
Solubility 3.76e-03 mg/ml ; 6.92e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.37
Solubility 2.33e-05 mg/ml ; 4.29e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.14 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.02
Molecule 50
SMILES CCOC(=O)c1ccc(cc1)NC(=O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C28H18N3O9S
Molecular weight 572.52 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 10
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 146.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

220.13 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.28
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.36
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.51
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.29
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.94
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.76
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.02
Solubility 5.47e-02 mg/ml ; 9.56e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.59
Solubility 1.49e-03 mg/ml ; 2.60e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.44
Solubility 2.08e-06 mg/ml ; 3.63e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.83 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.28
Molecule 51
SMILES O=C(Nc1cccc2c1cccc2)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C29H16N3O7S
Molecular weight 550.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 28
Fraction Csp3 0.03
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 148.34
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.26
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.49
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.51
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.50
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.43
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.82
Solubility 8.31e-03 mg/ml ; 1.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.11
Solubility 4.25e-04 mg/ml ; 7.72e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.66
Solubility 1.20e-07 mg/ml ; 2.18e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.17
Molecule 52
SMILES O=C(Nc1ccc2c(c1)cccc2)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C29H16N3O7S
Molecular weight 550.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 28
Fraction Csp3 0.03
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 148.34
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.79
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.49
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.51
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.50
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.34
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.82
Solubility 8.31e-03 mg/ml ; 1.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.11
Solubility 4.25e-04 mg/ml ; 7.72e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.66
Solubility 1.20e-07 mg/ml ; 2.18e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.17
Molecule 53
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCN(CC1)c1ccccc1
Physicochemical Properties
Formula C26H19N2O7
Molecular weight 471.44 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 132.56
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.73
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.50
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.38
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.75
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.42
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.89
Solubility 6.08e+00 mg/ml ; 1.29e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.83
Solubility 6.94e+01 mg/ml ; 1.47e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.34
Solubility 2.14e-04 mg/ml ; 4.55e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.24 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 54
SMILES Cc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C27H21N2O7
Molecular weight 485.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 137.53
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.14
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.14
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.07
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.55
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.96
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.99
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.19
Solubility 3.12e+00 mg/ml ; 6.42e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.21
Solubility 3.02e+01 mg/ml ; 6.23e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.71
Solubility 9.36e-05 mg/ml ; 1.93e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.07 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_E
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.07
Molecule 55
SMILES Clc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H18ClN2O7
Molecular weight 505.88 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 137.57
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.30
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.88
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.72
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.07
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.82
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.48
Solubility 1.67e+00 mg/ml ; 3.29e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.48
Solubility 1.69e+01 mg/ml ; 3.35e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.92
Solubility 6.08e-05 mg/ml ; 1.20e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.01 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.94
Molecule 56
SMILES Brc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H18BrN2O7
Molecular weight 550.33 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 140.26
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.31
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.81
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.61
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.18
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.11
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.76
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.80
Solubility 8.68e-01 mg/ml ; 1.58e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.55
Solubility 1.56e+01 mg/ml ; 2.83e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.10
Solubility 4.34e-05 mg/ml ; 7.89e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.23 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.97
Molecule 57
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCN(CC1)c1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C27H18Cl3N2O7
Molecular weight 588.80 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 151.80
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.72
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.34
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.70
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.18
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.05
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.53
Solubility 1.73e-01 mg/ml ; 2.95e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.51
Solubility 1.81e+00 mg/ml ; 3.07e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.10
Solubility 4.69e-06 mg/ml ; 7.96e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.81 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.07
Molecule 58
SMILES Oc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C26H19N2O8
Molecular weight 487.44 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 5
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 134.58
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

155.35 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-4.12
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.86
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.67
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.25
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.95
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.79
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.75
Solubility 8.65e+00 mg/ml ; 1.77e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.88
Solubility 6.38e+01 mg/ml ; 1.31e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.75
Solubility 8.67e-04 mg/ml ; 1.78e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.59 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 59
SMILES COc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C27H21N2O8
Molecular weight 501.46 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 9
Num. H-bond donors 1
Molar Refractivity 139.05
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

144.35 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.19
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.37
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.05
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.50
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.33
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.97
Solubility 5.38e+00 mg/ml ; 1.07e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.99
Solubility 5.07e+01 mg/ml ; 1.01e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.44
Solubility 1.83e-04 mg/ml ; 3.66e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.45 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.07
Molecule 60
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCN(CC1)c1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C26H19N3O9
Molecular weight 517.44 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 6
Num. H-bond acceptors 10
Num. H-bond donors 2
Molar Refractivity 137.64
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

175.43 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.14
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.58
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.74
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.28
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.24
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.67
Solubility 1.10e+01 mg/ml ; 2.12e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.01
Solubility 5.00e+01 mg/ml ; 9.67e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.67
Solubility 1.10e-03 mg/ml ; 2.13e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.98 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: XLOGP3<-2, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.07
Molecule 61
SMILES [O-]C(=O)c1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C27H18N2O9
Molecular weight 514.44 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 6
Num. H-bond acceptors 10
Num. H-bond donors 1
Molar Refractivity 137.58
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

175.25 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-11.62
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.98
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-3.01
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.07
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.90
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-3.56
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.76
Solubility 9.02e+00 mg/ml ; 1.75e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.18
Solubility 3.42e+01 mg/ml ; 6.66e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.67
Solubility 1.11e-03 mg/ml ; 2.15e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.84 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.03
Molecule 62
SMILES CCOC(=O)c1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-]
Physicochemical Properties
Formula C29H23N2O9
Molecular weight 543.50 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 19
Fraction Csp3 0.24
Num. rotatable bonds 8
Num. H-bond acceptors 10
Num. H-bond donors 1
Molar Refractivity 148.65
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

161.42 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.20
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.28
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.20
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.67
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.87
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.10
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.23
Solubility 3.22e+00 mg/ml ; 5.93e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.61
Solubility 1.33e+01 mg/ml ; 2.44e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.74
Solubility 9.87e-05 mg/ml ; 1.82e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.52 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.30
Molecule 63
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCN(CC1)c1cccc2c1cccc2
Physicochemical Properties
Formula C30H21N2O7
Molecular weight 521.50 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 23
Fraction Csp3 0.17
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 150.07
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.14
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.25
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.22
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.44
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.46
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.02
Solubility 4.96e-01 mg/ml ; 9.50e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.13
Solubility 3.87e+00 mg/ml ; 7.43e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.96
Solubility 5.69e-06 mg/ml ; 1.09e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.66 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 64
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)[O-])C[N+]1CCN(CC1)c1ccc2c(c1)cccc2
Physicochemical Properties
Formula C30H21N2O7
Molecular weight 521.50 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 23
Fraction Csp3 0.17
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 150.07
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.42
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.25
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.22
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.44
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.52
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.02
Solubility 4.96e-01 mg/ml ; 9.50e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.13
Solubility 3.87e+00 mg/ml ; 7.43e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.96
Solubility 5.69e-06 mg/ml ; 1.09e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.66 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 65
SMILES O=c1ccc(c2c1cc1=c2c(=O)ccc1=O)O
Physicochemical Properties
Formula C13H6O4
Molecular weight 226.18 g/mol
Num. heavy atoms 17
Num. arom. heavy atoms 13
Fraction Csp3 0.00
Num. rotatable bonds 0
Num. H-bond acceptors 4
Num. H-bond donors 1
Molar Refractivity 61.57
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

71.44 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.40
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.02
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.18
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.05
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.07
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.92
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.80
Solubility 3.62e+00 mg/ml ; 1.60e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.03
Solubility 2.11e+01 mg/ml ; 9.32e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.47
Solubility 7.63e-03 mg/ml ; 3.37e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.69 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW<250
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.12
Molecule 66
SMILES SC(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C15H8O5S
Molecular weight 300.29 g/mol
Num. heavy atoms 21
Num. arom. heavy atoms 13
Fraction Csp3 0.07
Num. rotatable bonds 2
Num. H-bond acceptors 5
Num. H-bond donors 1
Molar Refractivity 79.47
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

127.31 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.64
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.05
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.18
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.43
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.61
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.99
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.00
Solubility 3.03e+00 mg/ml ; 1.01e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.17
Solubility 2.02e+00 mg/ml ; 6.72e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.92
Solubility 3.63e-03 mg/ml ; 1.21e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.17 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

2 alerts: thioester, thiol_2

Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.23
Molecule 67
SMILES NC(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C15H9NO5
Molecular weight 283.24 g/mol
Num. heavy atoms 21
Num. arom. heavy atoms 13
Fraction Csp3 0.07
Num. rotatable bonds 2
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 74.25
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

114.53 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.98
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.20
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.79
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.84
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.61
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.15
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.17
Solubility 1.93e+01 mg/ml ; 6.82e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.71
Solubility 5.51e+01 mg/ml ; 1.94e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.46
Solubility 9.93e-03 mg/ml ; 3.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.88 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.17
Molecule 68
SMILES [N+]CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C15H7NO5
Molecular weight 281.22 g/mol
Num. heavy atoms 21
Num. arom. heavy atoms 13
Fraction Csp3 0.07
Num. rotatable bonds 2
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 72.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

88.51 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.04
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.57
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.93
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.61
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.78
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.25
Solubility 1.57e+01 mg/ml ; 5.57e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.33
Solubility 1.31e+02 mg/ml ; 4.67e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.46
Solubility 9.86e-03 mg/ml ; 3.51e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.75 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.08
Molecule 69
SMILES CCCC[N+]CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C19H16NO5
Molecular weight 338.33 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.26
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 93.13
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

88.51 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.45
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.72
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.77
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.96
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.24
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.14
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.38
Solubility 1.41e+00 mg/ml ; 4.17e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.16
Solubility 2.36e+00 mg/ml ; 6.97e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.44
Solubility 1.23e-04 mg/ml ; 3.64e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.85 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 70
SMILES OCC[N+]CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C17H12NO6
Molecular weight 326.28 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.18
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 84.67
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

108.74 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.21
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.23
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.85
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.55
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.17
Solubility 2.20e+01 mg/ml ; 6.74e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.58
Solubility 8.60e+01 mg/ml ; 2.64e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.08
Solubility 2.72e-03 mg/ml ; 8.34e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.15 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.30
Molecule 71
SMILES C[N+](CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C
Physicochemical Properties
Formula C17H13NO5
Molecular weight 311.29 g/mol
Num. heavy atoms 23
Num. arom. heavy atoms 13
Fraction Csp3 0.18
Num. rotatable bonds 3
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 84.57
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

91.75 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.96
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.06
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.24
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.43
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.00
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.05
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.95
Solubility 3.47e+00 mg/ml ; 1.12e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.42
Solubility 1.20e+01 mg/ml ; 3.84e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.91
Solubility 3.80e-03 mg/ml ; 1.22e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.24 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.32
Molecule 72
SMILES CC[N+](CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CC
Physicochemical Properties
Formula C19H17NO5
Molecular weight 339.34 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.26
Num. rotatable bonds 5
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 94.18
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

91.75 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.30
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.67
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.54
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.96
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.79
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.67
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.42
Solubility 1.29e+00 mg/ml ; 3.79e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.17
Solubility 2.28e+00 mg/ml ; 6.71e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.71
Solubility 6.68e-04 mg/ml ; 1.97e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.89 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.52
Molecule 73
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCCCC1
Physicochemical Properties
Formula C20H17NO5
Molecular weight 351.35 g/mol
Num. heavy atoms 26
Num. arom. heavy atoms 13
Fraction Csp3 0.30
Num. rotatable bonds 3
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 100.50
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

91.75 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.79
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.78
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.30
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.74
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.93
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.61
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.68
Solubility 7.31e-01 mg/ml ; 2.08e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.29
Solubility 1.81e+00 mg/ml ; 5.16e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.50
Solubility 1.11e-03 mg/ml ; 3.15e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.89 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.53
Molecule 74
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCOCC1
Physicochemical Properties
Formula C19H15NO6
Molecular weight 353.33 g/mol
Num. heavy atoms 26
Num. arom. heavy atoms 13
Fraction Csp3 0.26
Num. rotatable bonds 3
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 96.77
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.98 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.47
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.43
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.85
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.76
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.29
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.26
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.93
Solubility 4.13e+00 mg/ml ; 1.17e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.23
Solubility 2.10e+01 mg/ml ; 5.95e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.96
Solubility 3.86e-03 mg/ml ; 1.09e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.76 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 75
SMILES O=C1C(=C(C1=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)O
Physicochemical Properties
Formula C17H6O7
Molecular weight 322.23 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.00
Num. rotatable bonds 1
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 81.13
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

125.81 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.74
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.87
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.40
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.69
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.12
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.18
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.62
Solubility 7.65e+00 mg/ml ; 2.37e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.29
Solubility 1.65e+01 mg/ml ; 5.12e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.72
Solubility 6.08e-03 mg/ml ; 1.89e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.88 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.43
Molecule 76
SMILES O=C1C(=C(C1=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)O
Physicochemical Properties
Formula C18H8O7
Molecular weight 336.25 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 2
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 85.15
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

125.81 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.84
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.58
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.32
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.45
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.51
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.40
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.81
Solubility 5.18e+00 mg/ml ; 1.54e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.59
Solubility 8.62e+00 mg/ml ; 2.56e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.12
Solubility 2.54e-03 mg/ml ; 7.57e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.76 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.45
Molecule 77
SMILES O=C1NC(=O)C(S1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C16H7NO6S
Molecular weight 341.29 g/mol
Num. heavy atoms 24
Num. arom. heavy atoms 13
Fraction Csp3 0.06
Num. rotatable bonds 1
Num. H-bond acceptors 6
Num. H-bond donors 2
Molar Refractivity 89.13
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

142.91 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.88
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.17
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.50
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.65
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.87
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.48
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.18
Solubility 2.24e+00 mg/ml ; 6.55e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.38
Solubility 1.44e+00 mg/ml ; 4.21e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.97
Solubility 3.66e-03 mg/ml ; 1.07e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.50 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.77
Molecule 78
SMILES O=C1NC(=O)C(S1)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C17H9NO6S
Molecular weight 355.32 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.12
Num. rotatable bonds 2
Num. H-bond acceptors 6
Num. H-bond donors 2
Molar Refractivity 93.94
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

142.91 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.17
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.13
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.31
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.40
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.24
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.77
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.38
Solubility 1.49e+00 mg/ml ; 4.19e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.69
Solubility 7.30e-01 mg/ml ; 2.05e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.37
Solubility 1.53e-03 mg/ml ; 4.31e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.38 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.82
Molecule 79
SMILES Oc1ccc(=O)c2c1c1=c(c2)c(=O)cc(c1=O)c1ncsc1
Physicochemical Properties
Formula C16H7NO4S
Molecular weight 309.30 g/mol
Num. heavy atoms 22
Num. arom. heavy atoms 18
Fraction Csp3 0.00
Num. rotatable bonds 1
Num. H-bond acceptors 5
Num. H-bond donors 1
Molar Refractivity 82.68
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

112.57 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.97
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.68
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.30
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.29
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.72
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.68
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.73
Solubility 5.82e-01 mg/ml ; 1.88e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.62
Solubility 7.41e-01 mg/ml ; 2.39e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.86
Solubility 4.24e-04 mg/ml ; 1.37e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.70 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.23
Molecule 80
SMILES Oc1ccc(=O)c2c1c1=c(c2)c(=O)cc(c1=O)Cc1cscn1
Physicochemical Properties
Formula C17H9NO4S
Molecular weight 323.32 g/mol
Num. heavy atoms 23
Num. arom. heavy atoms 18
Fraction Csp3 0.06
Num. rotatable bonds 2
Num. H-bond acceptors 5
Num. H-bond donors 1
Molar Refractivity 86.69
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

112.57 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.05
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.97
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.23
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.04
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.09
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.86
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.90
Solubility 4.04e-01 mg/ml ; 1.25e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.92
Solubility 3.87e-01 mg/ml ; 1.20e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.26
Solubility 1.78e-04 mg/ml ; 5.50e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.58 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.31
Molecule 81
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccccc1
Physicochemical Properties
Formula C21H13NO5
Molecular weight 359.33 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 100.59
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.66
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.58
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.28
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.23
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.16
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.78
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.32
Solubility 1.72e-01 mg/ml ; 4.79e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.30
Solubility 1.79e-01 mg/ml ; 4.99e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.34
Solubility 1.64e-05 mg/ml ; 4.56e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.37 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.35
Molecule 82
SMILES Cc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C22H15NO5
Molecular weight 373.36 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 105.56
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.80
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.95
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.59
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.45
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.68
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.09
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.62
Solubility 8.93e-02 mg/ml ; 2.39e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.69
Solubility 7.69e-02 mg/ml ; 2.06e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.72
Solubility 7.16e-06 mg/ml ; 1.92e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.19 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 83
SMILES Clc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C21H12ClNO5
Molecular weight 393.78 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 105.60
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.83
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.21
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.94
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.72
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.80
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.30
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.91
Solubility 4.82e-02 mg/ml ; 1.23e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.96
Solubility 4.36e-02 mg/ml ; 1.11e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.93
Solubility 4.63e-06 mg/ml ; 1.17e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.13 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.33
Molecule 84
SMILES Brc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C21H12BrNO5
Molecular weight 438.23 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 108.29
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.98
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.27
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.82
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.84
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.23
Solubility 2.61e-02 mg/ml ; 5.95e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.02
Solubility 4.20e-02 mg/ml ; 9.59e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.12
Solubility 3.29e-06 mg/ml ; 7.50e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.36 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.37
Molecule 85
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C22H12Cl3NO5
Molecular weight 476.69 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 5
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 119.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.18
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.20
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.00
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

1.35
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.90
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.13
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.94
Solubility 5.54e-03 mg/ml ; 1.16e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.98
Solubility 4.96e-03 mg/ml ; 1.04e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.13
Solubility 3.53e-07 mg/ml ; 7.41e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-6.94 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 86
SMILES Oc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C21H13NO6
Molecular weight 375.33 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 102.61
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

120.77 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.15
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.23
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.99
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.30
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.68
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.35
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.18
Solubility 2.48e-01 mg/ml ; 6.61e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.36
Solubility 1.62e-01 mg/ml ; 4.33e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.75
Solubility 6.61e-05 mg/ml ; 1.76e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.72 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.39
Molecule 87
SMILES COc1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C22H15NO6
Molecular weight 389.36 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 5
Num. H-bond acceptors 6
Num. H-bond donors 2
Molar Refractivity 107.08
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

109.77 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.72
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.55
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.29
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.08
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.22
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.74
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.39
Solubility 1.60e-01 mg/ml ; 4.12e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.46
Solubility 1.34e-01 mg/ml ; 3.43e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.45
Solubility 1.40e-05 mg/ml ; 3.59e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.57 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 88
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C21H13N2O7
Molecular weight 405.34 g/mol
Num. heavy atoms 30
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 3
Molar Refractivity 105.66
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.85 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.95
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.08
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.16
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.00
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.84
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.09
Solubility 3.29e-01 mg/ml ; 8.12e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.49
Solubility 1.30e-01 mg/ml ; 3.20e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.68
Solubility 8.37e-05 mg/ml ; 2.07e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.10 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 89
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C22H12NO7
Molecular weight 402.33 g/mol
Num. heavy atoms 30
Num. arom. heavy atoms 19
Fraction Csp3 0.05
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 105.61
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.22
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.11
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.35
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.23
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.62
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.87
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.17
Solubility 2.70e-01 mg/ml ; 6.72e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.66
Solubility 8.86e-02 mg/ml ; 2.20e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.68
Solubility 8.41e-05 mg/ml ; 2.09e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.97 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.41
Molecule 90
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H17NO7
Molecular weight 431.39 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.12
Num. rotatable bonds 7
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 116.68
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

126.84 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.84
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.80
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.46
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.30
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.57
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.99
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.63
Solubility 1.02e-01 mg/ml ; 2.37e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.08
Solubility 3.57e-02 mg/ml ; 8.27e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.76
Solubility 7.48e-06 mg/ml ; 1.73e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.65 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.67
Molecule 91
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1cccc2c1cccc2
Physicochemical Properties
Formula C25H15NO5
Molecular weight 409.39 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 23
Fraction Csp3 0.04
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 118.10
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.11
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.83
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.91
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.19
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.70
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.45
Solubility 1.47e-02 mg/ml ; 3.58e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.60
Solubility 1.03e-02 mg/ml ; 2.52e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.98
Solubility 4.33e-07 mg/ml ; 1.06e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-6.79 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.56
Molecule 92
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C25H15NO5
Molecular weight 409.39 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 23
Fraction Csp3 0.04
Num. rotatable bonds 4
Num. H-bond acceptors 5
Num. H-bond donors 2
Molar Refractivity 118.10
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

100.54 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.92
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.83
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.91
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.19
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.66
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.45
Solubility 1.47e-02 mg/ml ; 3.58e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.60
Solubility 1.03e-02 mg/ml ; 2.52e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.98
Solubility 4.33e-07 mg/ml ; 1.06e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-6.79 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.56
Molecule 93
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccccc1
Physicochemical Properties
Formula C24H15N3O5S
Molecular weight 457.46 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 125.82
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.65
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.90
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.97
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.20
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.01
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.15
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.02
Solubility 4.41e-02 mg/ml ; 9.65e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.75
Solubility 8.12e-03 mg/ml ; 1.78e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.72
Solubility 8.72e-07 mg/ml ; 1.91e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.74 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.81
Molecule 94
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)C
Physicochemical Properties
Formula C25H17N3O5S
Molecular weight 471.48 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 130.78
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.29
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.26
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.28
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.41
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.56
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.31
Solubility 2.29e-02 mg/ml ; 4.86e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.12
Solubility 3.54e-03 mg/ml ; 7.51e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.09
Solubility 3.81e-07 mg/ml ; 8.09e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.57 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.91
Molecule 95
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)Cl
Physicochemical Properties
Formula C24H14ClN3O5S
Molecular weight 491.90 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.96
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.62
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.67
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.66
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.69
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.61
Solubility 1.21e-02 mg/ml ; 2.45e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.40
Solubility 1.94e-03 mg/ml ; 3.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.30
Solubility 2.47e-07 mg/ml ; 5.02e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.50 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.78
Molecule 96
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)Br
Physicochemical Properties
Formula C24H14BrN3O5S
Molecular weight 536.35 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 133.52
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.95
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.59
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.73
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.78
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.75
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.92
Solubility 6.40e-03 mg/ml ; 1.19e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.47
Solubility 1.83e-03 mg/ml ; 3.42e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.48
Solubility 1.76e-07 mg/ml ; 3.29e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.73 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.82
Molecule 97
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C25H14Cl3N3O5S
Molecular weight 574.82 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 145.05
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.74
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.52
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.69
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

1.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.77
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.60
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.64
Solubility 1.32e-03 mg/ml ; 2.29e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.43
Solubility 2.13e-04 mg/ml ; 3.70e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.48
Solubility 1.90e-08 mg/ml ; 3.31e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.31 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, XLOGP3>3.5
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.92
Molecule 98
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)O
Physicochemical Properties
Formula C24H15N3O6S
Molecular weight 473.46 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 4
Molar Refractivity 127.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

173.93 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.76
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.68
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.31
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.84
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.87
Solubility 6.36e-02 mg/ml ; 1.34e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.80
Solubility 7.47e-03 mg/ml ; 1.58e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.13
Solubility 3.53e-06 mg/ml ; 7.45e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.09 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.85
Molecule 99
SMILES COc1ccc(cc1)NCC(=O)Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H17N3O6S
Molecular weight 487.48 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 7
Num. H-bond donors 3
Molar Refractivity 132.31
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

162.93 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.87
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.98
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.10
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.09
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.28
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.09
Solubility 4.00e-02 mg/ml ; 8.20e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.91
Solubility 5.95e-03 mg/ml ; 1.22e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.82
Solubility 7.45e-07 mg/ml ; 1.53e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.94
Molecule 100
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C24H15N4O7S
Molecular weight 503.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 4
Molar Refractivity 130.89
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

194.01 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.26
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.77
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.87
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.22
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.79
Solubility 8.23e-02 mg/ml ; 1.63e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.93
Solubility 5.88e-03 mg/ml ; 1.17e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.48e-06 mg/ml ; 8.89e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.48 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.92
Molecule 101
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C25H14N3O7S
Molecular weight 500.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.67
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.43
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.23
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.34
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.88
Solubility 6.67e-02 mg/ml ; 1.33e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.11
Solubility 3.92e-03 mg/ml ; 7.84e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.50e-06 mg/ml ; 8.99e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.34 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.84
Molecule 102
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C27H19N3O7S
Molecular weight 529.52 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 141.90
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.00 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.91
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.15
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.44
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.38
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.33
Solubility 2.47e-02 mg/ml ; 4.66e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.53
Solubility 1.56e-03 mg/ml ; 2.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.12
Solubility 4.01e-07 mg/ml ; 7.57e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.02 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.10
Molecule 103
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1cccc2c1cccc2
Physicochemical Properties
Formula C28H17N3O5S
Molecular weight 507.52 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 28
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 143.32
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.82
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.12
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.85
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.03
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.99
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.14
Solubility 3.72e-03 mg/ml ; 7.33e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.05
Solubility 4.55e-04 mg/ml ; 8.96e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.34
Solubility 2.31e-08 mg/ml ; 4.56e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.03
Molecule 104
SMILES O=C(Nc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C28H17N3O5S
Molecular weight 507.52 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 28
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 143.32
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.13
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.12
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.85
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.03
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.06
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.14
Solubility 3.72e-03 mg/ml ; 7.33e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.05
Solubility 4.55e-04 mg/ml ; 8.96e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.34
Solubility 2.31e-08 mg/ml ; 4.56e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.03
Molecule 105
SMILES O=C(Nc1ccccc1)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H15N3O5S
Molecular weight 457.46 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 125.82
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.47
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.90
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.97
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.20
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.01
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.31
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.02
Solubility 4.41e-02 mg/ml ; 9.65e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.75
Solubility 8.12e-03 mg/ml ; 1.78e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.72
Solubility 8.72e-07 mg/ml ; 1.91e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.74 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.84
Molecule 106
SMILES O=C(Nc1ccc(cc1)C)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H17N3O5S
Molecular weight 471.48 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 130.78
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.51
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.26
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.28
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.41
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.60
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.31
Solubility 2.29e-02 mg/ml ; 4.86e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.12
Solubility 3.54e-03 mg/ml ; 7.51e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.09
Solubility 3.81e-07 mg/ml ; 8.09e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.57 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.94
Molecule 107
SMILES O=C(Nc1ccc(cc1)Cl)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H14ClN3O5S
Molecular weight 491.90 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.86
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.62
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.67
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.66
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.87
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.61
Solubility 1.21e-02 mg/ml ; 2.45e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.40
Solubility 1.94e-03 mg/ml ; 3.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.30
Solubility 2.47e-07 mg/ml ; 5.02e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.50 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.81
Molecule 108
SMILES O=C(Nc1ccc(cc1)Br)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H14BrN3O5S
Molecular weight 536.35 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 133.52
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.63
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.59
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.73
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.78
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.89
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.92
Solubility 6.40e-03 mg/ml ; 1.19e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.47
Solubility 1.83e-03 mg/ml ; 3.42e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.48
Solubility 1.76e-07 mg/ml ; 3.29e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.73 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.85
Molecule 109
SMILES O=C(Nc1ccc(cc1)C(Cl)(Cl)Cl)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H14Cl3N3O5S
Molecular weight 574.82 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 145.05
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.97
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.52
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.69
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

1.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.77
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.64
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.64
Solubility 1.32e-03 mg/ml ; 2.29e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.43
Solubility 2.13e-04 mg/ml ; 3.70e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.48
Solubility 1.90e-08 mg/ml ; 3.31e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.31 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, XLOGP3>3.5
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 110
SMILES O=C(Nc1ccc(cc1)O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H15N3O6S
Molecular weight 473.46 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 4
Molar Refractivity 127.84
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

173.93 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.47
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.68
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.31
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.98
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.87
Solubility 6.36e-02 mg/ml ; 1.34e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.80
Solubility 7.47e-03 mg/ml ; 1.58e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.13
Solubility 3.53e-06 mg/ml ; 7.45e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.09 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.88
Molecule 111
SMILES COc1ccc(cc1)NC(=O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H17N3O6S
Molecular weight 487.48 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 7
Num. H-bond donors 3
Molar Refractivity 132.31
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

162.93 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.93
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.87
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.98
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.10
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.09
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.15
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.09
Solubility 4.00e-02 mg/ml ; 8.20e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.91
Solubility 5.95e-03 mg/ml ; 1.22e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.82
Solubility 7.45e-07 mg/ml ; 1.53e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.95 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.97
Molecule 112
SMILES O=C(Nc1ccc(cc1)[N+](=O)O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C24H15N4O7S
Molecular weight 503.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 4
Molar Refractivity 130.89
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

194.01 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.26
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.77
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.87
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.22
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.79
Solubility 8.23e-02 mg/ml ; 1.63e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.93
Solubility 5.88e-03 mg/ml ; 1.17e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.48e-06 mg/ml ; 8.89e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.48 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 113
SMILES O=C(Nc1ccc(cc1)C(=O)[O-])CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H14N3O7S
Molecular weight 500.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 24
Fraction Csp3 0.04
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 130.83
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.91
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.43
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.33
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.13
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.61
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.88
Solubility 6.67e-02 mg/ml ; 1.33e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.11
Solubility 3.92e-03 mg/ml ; 7.84e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.05
Solubility 4.50e-06 mg/ml ; 8.99e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.34 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.87
Molecule 114
SMILES CCOC(=O)c1ccc(cc1)NC(=O)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C27H19N3O7S
Molecular weight 529.52 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 141.90
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.00 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.90
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

2.15
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.44
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.58
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.33
Solubility 2.47e-02 mg/ml ; 4.66e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.53
Solubility 1.56e-03 mg/ml ; 2.94e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.12
Solubility 4.01e-07 mg/ml ; 7.57e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.02 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.13
Molecule 115
SMILES O=C(Nc1cccc2c1cccc2)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C28H17N3O5S
Molecular weight 507.52 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 28
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 143.32
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

3.11
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.12
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.85
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.03
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.25
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.14
Solubility 3.72e-03 mg/ml ; 7.33e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.05
Solubility 4.55e-04 mg/ml ; 8.96e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.34
Solubility 2.31e-08 mg/ml ; 4.56e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 116
SMILES O=C(Nc1ccc2c(c1)cccc2)CNc1scc(n1)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C28H17N3O5S
Molecular weight 507.52 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 28
Fraction Csp3 0.04
Num. rotatable bonds 6
Num. H-bond acceptors 6
Num. H-bond donors 3
Molar Refractivity 143.32
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

153.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.54
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

3.12
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.85
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.03
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

3.14
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.14
Solubility 3.72e-03 mg/ml ; 7.33e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.05
Solubility 4.55e-04 mg/ml ; 8.96e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.34
Solubility 2.31e-08 mg/ml ; 4.56e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.16 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 117
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1ccccc1
Physicochemical Properties
Formula C25H20N2O5
Molecular weight 428.44 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.20
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 127.54
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.97
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.26
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.40
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.96
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.06
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.64
Solubility 9.92e-02 mg/ml ; 2.31e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.13
Solubility 3.15e-01 mg/ml ; 7.35e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.01
Solubility 4.16e-05 mg/ml ; 9.71e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.83 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.77
Molecule 118
SMILES Cc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C26H22N2O5
Molecular weight 442.46 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.23
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 132.51
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.79
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.90
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.57
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.20
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.49
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.31
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.94
Solubility 5.05e-02 mg/ml ; 1.14e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.52
Solubility 1.34e-01 mg/ml ; 3.04e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.39
Solubility 1.82e-05 mg/ml ; 4.10e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.65 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_E
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.88
Molecule 119
SMILES Clc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H19ClN2O5
Molecular weight 462.88 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.20
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 132.55
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.53
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.91
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.93
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.60
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.45
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.23
Solubility 2.71e-02 mg/ml ; 5.85e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.79
Solubility 7.55e-02 mg/ml ; 1.63e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.59
Solubility 1.18e-05 mg/ml ; 2.54e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.59 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.75
Molecule 120
SMILES Brc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H19BrN2O5
Molecular weight 507.33 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.20
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 135.24
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.63
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.22
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.02
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.83
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.64
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.54
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.55
Solubility 1.44e-02 mg/ml ; 2.84e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.85
Solubility 7.17e-02 mg/ml ; 1.41e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.78
Solubility 8.38e-06 mg/ml ; 1.65e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.82 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.79
Molecule 121
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C26H19Cl3N2O5
Molecular weight 545.80 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.23
Num. rotatable bonds 5
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 146.78
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

3.48
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

3.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.98
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.33
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.40
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.27
Solubility 2.94e-03 mg/ml ; 5.38e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.81
Solubility 8.36e-03 mg/ml ; 1.53e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.78
Solubility 9.04e-07 mg/ml ; 1.66e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.39 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.89
Molecule 122
SMILES Oc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C25H20N2O6
Molecular weight 444.44 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.20
Num. rotatable bonds 4
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 129.57
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

115.22 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.89
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.18
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.04
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.92
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.48
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.52
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.50
Solubility 1.40e-01 mg/ml ; 3.16e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.20
Solubility 2.83e-01 mg/ml ; 6.38e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.42
Solubility 1.68e-04 mg/ml ; 3.78e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.17 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.81
Molecule 123
SMILES COc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C26H22N2O6
Molecular weight 458.46 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 19
Fraction Csp3 0.23
Num. rotatable bonds 5
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 134.04
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

104.22 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.42
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.50
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.27
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.72
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.03
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.90
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.71
Solubility 8.92e-02 mg/ml ; 1.95e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.30
Solubility 2.32e-01 mg/ml ; 5.05e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.11
Solubility 3.55e-05 mg/ml ; 7.74e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.03 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.89
Molecule 124
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C25H20N3O7
Molecular weight 474.44 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 19
Fraction Csp3 0.20
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 132.62
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.30 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.90
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.06
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.80
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.14
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.42
Solubility 1.80e-01 mg/ml ; 3.80e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.33
Solubility 2.24e-01 mg/ml ; 4.72e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.35
Solubility 2.13e-04 mg/ml ; 4.50e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.56 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.89
Molecule 125
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C26H19N2O7
Molecular weight 471.44 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 19
Fraction Csp3 0.19
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 132.56
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.55
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.22
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.38
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.75
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.42
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.30
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.07
Solubility 4.05e+00 mg/ml ; 8.59e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.12
Solubility 3.55e+01 mg/ml ; 7.54e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.34
Solubility 2.14e-04 mg/ml ; 4.55e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.04 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.83
Molecule 126
SMILES CCOC(=O)c1ccc(cc1)N1CC[N+](CC1)CC(=O)c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C28H24N2O7
Molecular weight 500.50 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.25
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 143.63
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

121.29 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.33
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.75
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.35
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.38
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.11
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.96
Solubility 5.44e-02 mg/ml ; 1.09e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.91
Solubility 6.10e-02 mg/ml ; 1.22e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.42
Solubility 1.91e-05 mg/ml ; 3.81e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.11 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.10
Molecule 127
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1cccc2c1cccc2
Physicochemical Properties
Formula C29H22N2O5
Molecular weight 478.50 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 23
Fraction Csp3 0.17
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 145.05
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.72
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.78
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.41
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.77
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.97
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.82
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.77
Solubility 8.19e-03 mg/ml ; 1.71e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.43
Solubility 1.77e-02 mg/ml ; 3.71e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.64
Solubility 1.10e-06 mg/ml ; 2.30e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.25 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 128
SMILES O=C(c1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2)C[N+]1CCN(CC1)c1ccc2c(c1)cccc2
Physicochemical Properties
Formula C29H22N2O5
Molecular weight 478.50 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 23
Fraction Csp3 0.17
Num. rotatable bonds 4
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 145.05
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

94.99 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.39
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.78
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.41
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.77
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.97
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.76
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.77
Solubility 8.19e-03 mg/ml ; 1.71e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.43
Solubility 1.77e-02 mg/ml ; 3.71e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.64
Solubility 1.10e-06 mg/ml ; 2.30e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.25 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 129
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2
Physicochemical Properties
Formula C15H7O6
Molecular weight 283.21 g/mol
Num. heavy atoms 21
Num. arom. heavy atoms 13
Fraction Csp3 0.07
Num. rotatable bonds 2
Num. H-bond acceptors 6
Num. H-bond donors 1
Molar Refractivity 71.17
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

111.57 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.19
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.55
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.53
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.43
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.85
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.31
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.58
Solubility 7.53e+00 mg/ml ; 2.66e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.32
Solubility 1.34e+01 mg/ml ; 4.75e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.24
Solubility 1.63e-02 mg/ml ; 5.77e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.42 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.17
Molecule 130
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2CC(=O)S
Physicochemical Properties
Formula C17H9O7S
Molecular weight 357.31 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.12
Num. rotatable bonds 4
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 89.07
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

167.44 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.34
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.03
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.53
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.81
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.47
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.29
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.53
Solubility 1.06e+01 mg/ml ; 2.97e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.00
Solubility 3.58e+00 mg/ml ; 1.00e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.66
Solubility 7.78e-03 mg/ml ; 2.18e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

2 alerts: thioester, thiol_2

Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.42
Molecule 131
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2CC(=O)N
Physicochemical Properties
Formula C17H10NO7
Molecular weight 340.26 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.12
Num. rotatable bonds 4
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 83.85
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

154.66 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.66
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.19
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.50
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.47
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.55
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.69
Solubility 6.94e+01 mg/ml ; 2.04e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.53
Solubility 1.01e+02 mg/ml ; 2.98e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.20
Solubility 2.13e-02 mg/ml ; 6.25e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.93 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: XLOGP3<-2, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.39
Molecule 132
SMILES [N+]CC(=O)c1c2c(=O)ccc(c2c2=c1c(=O)cc(c2=O)CC(=O)[O-])O
Physicochemical Properties
Formula C17H8NO7
Molecular weight 338.25 g/mol
Num. heavy atoms 25
Num. arom. heavy atoms 13
Fraction Csp3 0.12
Num. rotatable bonds 4
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 82.44
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

128.64 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-4.34
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.28
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.47
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.89
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

0.68
Solubility 1.61e+03 mg/ml ; 4.75e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

2.25
Solubility 6.03e+04 mg/ml ; 1.78e+02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.20
Solubility 2.11e-02 mg/ml ; 6.25e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.44 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		Yes
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.33
Molecule 133
SMILES CCCC[N+]CC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C21H17NO7
Molecular weight 395.36 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.29
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 102.73
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

128.64 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-10.77
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.58
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.94
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.36
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.14
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-2.90
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.47
Solubility 1.34e+02 mg/ml ; 3.39e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.42
Solubility 1.05e+03 mg/ml ; 2.66e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.17
Solubility 2.65e-04 mg/ml ; 6.70e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.54 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		Yes
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.71
Molecule 134
SMILES OCC[N+]CC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C19H13NO8
Molecular weight 383.31 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.21
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 94.28
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

148.87 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-10.59
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-4.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.75
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.59
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.74
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-4.14
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

0.74
Solubility 2.12e+03 mg/ml ; 5.54e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

2.00
Solubility 3.86e+04 mg/ml ; 1.01e+02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.82
Solubility 5.84e-03 mg/ml ; 1.52e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.84 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.55
Molecule 135
SMILES C[N+](CC(=O)c1c2c(=O)ccc(c2c2=c1c(=O)cc(c2=O)CC(=O)[O-])O)C
Physicochemical Properties
Formula C19H14NO7
Molecular weight 368.32 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 13
Fraction Csp3 0.21
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 94.17
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.66
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-3.36
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.95
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.81
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.88
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-2.18
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.03
Solubility 3.41e+02 mg/ml ; 9.27e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

1.17
Solubility 5.40e+03 mg/ml ; 1.47e+01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.65
Solubility 8.17e-03 mg/ml ; 2.22e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.93 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.56
Molecule 136
SMILES CC[N+](CC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CC
Physicochemical Properties
Formula C21H18NO7
Molecular weight 396.37 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.29
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 103.79
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.22
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.63
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.17
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.36
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.69
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.54
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.51
Solubility 1.22e+02 mg/ml ; 3.09e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.41
Solubility 1.02e+03 mg/ml ; 2.56e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.44
Solubility 1.44e-03 mg/ml ; 3.62e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.59 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.77
Molecule 137
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCCCC1
Physicochemical Properties
Formula C22H18NO7
Molecular weight 408.38 g/mol
Num. heavy atoms 30
Num. arom. heavy atoms 13
Fraction Csp3 0.32
Num. rotatable bonds 5
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 110.10
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

131.88 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.25
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.51
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.40
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.14
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.78
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.51
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.78
Solubility 6.76e+01 mg/ml ; 1.65e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

0.28
Solubility 7.85e+02 mg/ml ; 1.92e+00 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.24
Solubility 2.38e-03 mg/ml ; 5.82e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.57 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.78
Molecule 138
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCOCC1
Physicochemical Properties
Formula C21H16NO8
Molecular weight 410.35 g/mol
Num. heavy atoms 30
Num. arom. heavy atoms 13
Fraction Csp3 0.29
Num. rotatable bonds 5
Num. H-bond acceptors 9
Num. H-bond donors 1
Molar Refractivity 106.38
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

141.11 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.74
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-3.73
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.56
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-5.14
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.14
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-2.21
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-0.02
Solubility 3.87e+02 mg/ml ; 9.44e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

1.36
Solubility 9.32e+03 mg/ml ; 2.27e+01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Highly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.70
Solubility 8.28e-03 mg/ml ; 2.02e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.45 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: XLOGP3<-2
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.72
Molecule 139
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C1=C(O)C(=O)C1=O
Physicochemical Properties
Formula C19H7O9
Molecular weight 379.25 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.05
Num. rotatable bonds 3
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 90.73
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

165.94 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-0.53
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.86
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.11
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.04
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.97
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.71
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.17
Solubility 2.59e+01 mg/ml ; 6.84e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.11
Solubility 2.97e+01 mg/ml ; 7.84e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.46
Solubility 1.30e-02 mg/ml ; 3.44e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.93 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.65
Molecule 140
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2CC1=C(O)C(=O)C1=O
Physicochemical Properties
Formula C20H9O9
Molecular weight 393.28 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.10
Num. rotatable bonds 4
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 94.75
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

165.94 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.78
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.57
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.02
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.82
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.37
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.25
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.36
Solubility 1.73e+01 mg/ml ; 4.40e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.41
Solubility 1.54e+01 mg/ml ; 3.92e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.86
Solubility 5.47e-03 mg/ml ; 1.39e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.81 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: imine_one_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: diketo_group
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.64
Molecule 141
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C1SC(=O)NC1=O
Physicochemical Properties
Formula C18H8NO8S
Molecular weight 398.32 g/mol
Num. heavy atoms 28
Num. arom. heavy atoms 13
Fraction Csp3 0.11
Num. rotatable bonds 3
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 98.73
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

183.04 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.83
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.21
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.99
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.70
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.17
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.72
Solubility 7.51e+00 mg/ml ; 1.89e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.19
Solubility 2.56e+00 mg/ml ; 6.44e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-4.71
Solubility 7.86e-03 mg/ml ; 1.97e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.55 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.99
Molecule 142
SMILES O=C1SC(C(=O)N1)Cc1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C19H10NO8S
Molecular weight 412.35 g/mol
Num. heavy atoms 29
Num. arom. heavy atoms 13
Fraction Csp3 0.16
Num. rotatable bonds 4
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 103.54
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

183.04 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.97
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.86
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.02
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.76
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.09
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.09
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.92
Solubility 4.93e+00 mg/ml ; 1.20e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.50
Solubility 1.30e+00 mg/ml ; 3.14e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.10
Solubility 3.29e-03 mg/ml ; 7.98e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.43 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: thioester
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.02
Molecule 143
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2c1cscn1
Physicochemical Properties
Formula C18H8NO6S
Molecular weight 366.32 g/mol
Num. heavy atoms 26
Num. arom. heavy atoms 18
Fraction Csp3 0.06
Num. rotatable bonds 3
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 92.28
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

152.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.51
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.31
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.41
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.68
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.53
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.93
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.23
Solubility 2.16e+00 mg/ml ; 5.89e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.44
Solubility 1.34e+00 mg/ml ; 3.66e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-5.61
Solubility 9.10e-04 mg/ml ; 2.48e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.75 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.46
Molecule 144
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2Cc1cscn1
Physicochemical Properties
Formula C19H10NO6S
Molecular weight 380.35 g/mol
Num. heavy atoms 27
Num. arom. heavy atoms 18
Fraction Csp3 0.11
Num. rotatable bonds 4
Num. H-bond acceptors 7
Num. H-bond donors 1
Molar Refractivity 96.30
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

152.70 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.55
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.01
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.48
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.92
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.11
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.42
Solubility 1.44e+00 mg/ml ; 3.79e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.75
Solubility 6.80e-01 mg/ml ; 1.79e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.00
Solubility 3.81e-04 mg/ml ; 1.00e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.63 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.49
Molecule 145
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccccc1
Physicochemical Properties
Formula C23H14NO7
Molecular weight 416.36 g/mol
Num. heavy atoms 31
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 110.19
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.87
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.56
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.43
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.18
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.02
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.17
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.83
Solubility 6.13e-01 mg/ml ; 1.47e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.09
Solubility 3.41e-01 mg/ml ; 8.20e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.07
Solubility 3.53e-05 mg/ml ; 8.47e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.44 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.62
Molecule 146
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)C
Physicochemical Properties
Formula C24H16NO7
Molecular weight 430.39 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.12
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 115.16
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.57
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.93
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.12
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.03
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.56
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.14
Solubility 3.13e-01 mg/ml ; 7.28e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.47
Solubility 1.46e-01 mg/ml ; 3.39e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.45
Solubility 1.54e-05 mg/ml ; 3.58e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.27 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.73
Molecule 147
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)Cl
Physicochemical Properties
Formula C23H13ClNO7
Molecular weight 450.80 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 115.20
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.60
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.19
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.23
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.30
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.67
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.60
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.43
Solubility 1.68e-01 mg/ml ; 3.73e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.74
Solubility 8.20e-02 mg/ml ; 1.82e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.66
Solubility 9.98e-06 mg/ml ; 2.21e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.60
Molecule 148
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)Br
Physicochemical Properties
Formula C23H13BrNO7
Molecular weight 495.26 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 117.89
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.81
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.25
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.34
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.40
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.71
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.70
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.74
Solubility 8.98e-02 mg/ml ; 1.81e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.80
Solubility 7.81e-02 mg/ml ; 1.58e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.84
Solubility 7.11e-06 mg/ml ; 1.44e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.43 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MW>480
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.64
Molecule 149
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C24H13Cl3NO7
Molecular weight 533.72 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 19
Fraction Csp3 0.12
Num. rotatable bonds 7
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 129.43
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.18
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.29
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.91
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.79
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.55
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.46
Solubility 1.84e-02 mg/ml ; 3.45e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.77
Solubility 9.12e-03 mg/ml ; 1.71e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.84
Solubility 7.66e-07 mg/ml ; 1.44e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.01 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MW>480
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.74
Molecule 150
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)O
Physicochemical Properties
Formula C23H14NO8
Molecular weight 432.36 g/mol
Num. heavy atoms 32
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 112.22
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

160.90 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.41
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.21
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.72
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.69
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.55
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.75
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.70
Solubility 8.70e-01 mg/ml ; 2.01e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.15
Solubility 3.07e-01 mg/ml ; 7.11e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.48
Solubility 1.43e-04 mg/ml ; 3.30e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.79 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.66
Molecule 151
SMILES COc1ccc(cc1)NCC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C24H16NO8
Molecular weight 446.39 g/mol
Num. heavy atoms 33
Num. arom. heavy atoms 19
Fraction Csp3 0.12
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 2
Molar Refractivity 116.68
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

149.90 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.24
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.53
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.42
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.48
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.10
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.00
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.91
Solubility 5.55e-01 mg/ml ; 1.24e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.25
Solubility 2.52e-01 mg/ml ; 5.64e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.17
Solubility 3.01e-05 mg/ml ; 6.75e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.65 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.74
Molecule 152
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C23H14N2O9
Molecular weight 462.37 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 19
Fraction Csp3 0.09
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 115.27
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.98 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.07
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.63
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.89
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.20
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.61
Solubility 1.12e+00 mg/ml ; 2.43e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.28
Solubility 2.43e-01 mg/ml ; 5.26e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.41
Solubility 1.81e-04 mg/ml ; 3.91e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.17 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.74
Molecule 153
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C24H13NO9
Molecular weight 459.36 g/mol
Num. heavy atoms 34
Num. arom. heavy atoms 19
Fraction Csp3 0.08
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 115.21
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

180.80 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.18
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.09
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-2.06
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.61
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.51
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.22
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.70
Solubility 9.24e-01 mg/ml ; 2.01e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.44
Solubility 1.66e-01 mg/ml ; 3.62e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.40
Solubility 1.82e-04 mg/ml ; 3.95e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.04 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: WLOGP<-0.4
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.69
Molecule 154
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C26H18NO9
Molecular weight 488.42 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.15
Num. rotatable bonds 9
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 126.28
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

166.97 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.26
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.78
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.25
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.09
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.47
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.43
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.16
Solubility 3.41e-01 mg/ml ; 6.98e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-3.87
Solubility 6.63e-02 mg/ml ; 1.36e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.48
Solubility 1.62e-05 mg/ml ; 3.31e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.73 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 1 violation: MW>480
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.95
Molecule 155
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1cccc2c1cccc2
Physicochemical Properties
Formula C27H16NO7
Molecular weight 466.42 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 23
Fraction Csp3 0.07
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 127.70
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.19
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.81
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.73
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.48
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.06
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.05
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.96
Solubility 5.09e-02 mg/ml ; 1.09e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.38
Solubility 1.93e-02 mg/ml ; 4.14e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.70
Solubility 9.35e-07 mg/ml ; 2.00e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.86 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.84
Molecule 156
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C27H16NO7
Molecular weight 466.42 g/mol
Num. heavy atoms 35
Num. arom. heavy atoms 23
Fraction Csp3 0.07
Num. rotatable bonds 6
Num. H-bond acceptors 7
Num. H-bond donors 2
Molar Refractivity 127.70
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

140.67 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.81
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.73
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.48
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.06
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.93
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.96
Solubility 5.09e-02 mg/ml ; 1.09e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.38
Solubility 1.93e-02 mg/ml ; 4.14e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.70
Solubility 9.35e-07 mg/ml ; 2.00e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		Yes
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-7.86 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		Yes
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.56
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		3.84
Molecule 157
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccccc1
Physicochemical Properties
Formula C26H16N3O7S
Molecular weight 514.49 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.42
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.80
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.91
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.26
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.89
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.54
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.43e-01 mg/ml ; 2.79e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.57
Solubility 1.40e-02 mg/ml ; 2.72e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.44
Solubility 1.89e-06 mg/ml ; 3.67e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.79 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.08
Molecule 158
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C
Physicochemical Properties
Formula C27H18N3O7S
Molecular weight 528.51 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 140.38
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.77
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.28
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.57
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.01
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.43
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.81
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.86
Solubility 7.25e-02 mg/ml ; 1.37e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.95
Solubility 5.93e-03 mg/ml ; 1.12e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.80
Solubility 8.28e-07 mg/ml ; 1.57e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.62 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 159
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)Cl
Physicochemical Properties
Formula C26H15ClN3O7S
Molecular weight 548.93 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 140.43
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.07
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.92
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.07
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.15
Solubility 3.86e-02 mg/ml ; 7.03e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.22
Solubility 3.31e-03 mg/ml ; 6.03e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.01
Solubility 5.38e-07 mg/ml ; 9.79e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.56 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 160
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)Br
Physicochemical Properties
Formula C26H15BrN3O7S
Molecular weight 593.38 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 143.12
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.33
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.60
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.03
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.38
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.58
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.18
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.47
Solubility 2.03e-02 mg/ml ; 3.42e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.28
Solubility 3.10e-03 mg/ml ; 5.22e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.19
Solubility 3.85e-07 mg/ml ; 6.49e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.78 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.08
Molecule 161
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C27H15Cl3N3O7S
Molecular weight 631.85 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 154.66
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.21
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.98
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.87
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.67
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.85
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.20
Solubility 4.02e-03 mg/ml ; 6.36e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.26
Solubility 3.49e-04 mg/ml ; 5.53e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.18
Solubility 4.16e-08 mg/ml ; 6.59e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.35 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: MW>600, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.19
Molecule 162
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)O
Physicochemical Properties
Formula C26H16N3O8S
Molecular weight 530.49 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 9
Num. H-bond donors 4
Molar Refractivity 137.44
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

214.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.17
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.56
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.03
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.68
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.42
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.09
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.42
Solubility 2.01e-01 mg/ml ; 3.79e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.63
Solubility 1.25e-02 mg/ml ; 2.36e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.84
Solubility 7.65e-06 mg/ml ; 1.44e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.14 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.12
Molecule 163
SMILES COc1ccc(cc1)NCC(=O)Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C27H18N3O8S
Molecular weight 544.51 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 141.91
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

203.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.86
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.88
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.27
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.48
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.98
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.50
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.63
Solubility 1.27e-01 mg/ml ; 2.33e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.73
Solubility 1.02e-02 mg/ml ; 1.87e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.53
Solubility 1.62e-06 mg/ml ; 2.97e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.00 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.22
Molecule 164
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C26H16N4O9S
Molecular weight 560.49 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 9
Num. H-bond acceptors 10
Num. H-bond donors 4
Molar Refractivity 140.49
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

234.14 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.28
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.06
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.15
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.76
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.59
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.34
Solubility 2.55e-01 mg/ml ; 4.56e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.76
Solubility 9.77e-03 mg/ml ; 1.74e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.76
Solubility 9.74e-06 mg/ml ; 1.74e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.52 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 165
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C27H15N3O9S
Molecular weight 557.49 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.07
Num. rotatable bonds 9
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 140.44
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

233.96 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.44
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.37
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.60
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.38
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.59
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.42
Solubility 2.10e-01 mg/ml ; 3.77e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.92
Solubility 6.69e-03 mg/ml ; 1.20e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.75
Solubility 9.80e-06 mg/ml ; 1.76e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.39 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.12
Molecule 166
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C29H20N3O9S
Molecular weight 586.55 g/mol
Num. heavy atoms 42
Num. arom. heavy atoms 24
Fraction Csp3 0.14
Num. rotatable bonds 11
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 151.51
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

220.13 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.03
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.13
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.10
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.35
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.77
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.89
Solubility 7.64e-02 mg/ml ; 1.30e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.35
Solubility 2.64e-03 mg/ml ; 4.50e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.83
Solubility 8.74e-07 mg/ml ; 1.49e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.08 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 2 violations: Rotors>10, TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.39
Molecule 167
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1cccc2c1cccc2
Physicochemical Properties
Formula C30H18N3O7S
Molecular weight 564.54 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 28
Fraction Csp3 0.07
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 152.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.87
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.42
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.91
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.36
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.68
Solubility 1.18e-02 mg/ml ; 2.10e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.86
Solubility 7.74e-04 mg/ml ; 1.37e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.05
Solubility 5.04e-08 mg/ml ; 8.93e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.31
Molecule 168
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C30H18N3O7S
Molecular weight 564.54 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 28
Fraction Csp3 0.07
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 152.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.42
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.91
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.68
Solubility 1.18e-02 mg/ml ; 2.10e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.86
Solubility 7.74e-04 mg/ml ; 1.37e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.05
Solubility 5.04e-08 mg/ml ; 8.93e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.31
Molecule 169
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccccc1
Physicochemical Properties
Formula C26H16N3O7S
Molecular weight 514.49 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 135.42
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.80
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.91
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.26
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.19
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.89
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.54
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.56
Solubility 1.43e-01 mg/ml ; 2.79e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.57
Solubility 1.40e-02 mg/ml ; 2.72e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.44
Solubility 1.89e-06 mg/ml ; 3.67e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.79 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.08
Molecule 170
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C
Physicochemical Properties
Formula C27H18N3O7S
Molecular weight 528.51 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 140.38
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.77
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.28
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.57
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.01
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.43
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.81
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.86
Solubility 7.25e-02 mg/ml ; 1.37e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.95
Solubility 5.93e-03 mg/ml ; 1.12e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.80
Solubility 8.28e-07 mg/ml ; 1.57e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.62 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 171
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)Cl
Physicochemical Properties
Formula C26H15ClN3O7S
Molecular weight 548.93 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 140.43
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.07
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.92
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.28
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.54
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.07
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.15
Solubility 3.86e-02 mg/ml ; 7.03e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.22
Solubility 3.31e-03 mg/ml ; 6.03e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.01
Solubility 5.38e-07 mg/ml ; 9.79e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.56 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 172
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)Br
Physicochemical Properties
Formula C26H15BrN3O7S
Molecular weight 593.38 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 143.12
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.33
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.60
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.03
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.38
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.58
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.18
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.47
Solubility 2.03e-02 mg/ml ; 3.42e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.28
Solubility 3.10e-03 mg/ml ; 5.22e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-9.19
Solubility 3.85e-07 mg/ml ; 6.49e-10 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.78 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.08
Molecule 173
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C27H15Cl3N3O7S
Molecular weight 631.85 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 154.66
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.21
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.54
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.98
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.87
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

6.67
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.85
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-5.20
Solubility 4.02e-03 mg/ml ; 6.36e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-6.26
Solubility 3.49e-04 mg/ml ; 5.53e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.18
Solubility 4.16e-08 mg/ml ; 6.59e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.35 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: MW>600, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.19
Molecule 174
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)O
Physicochemical Properties
Formula C26H16N3O8S
Molecular weight 530.49 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 8
Num. H-bond acceptors 9
Num. H-bond donors 4
Molar Refractivity 137.44
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

214.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.17
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.56
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.03
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.68
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.42
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.09
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.42
Solubility 2.01e-01 mg/ml ; 3.79e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.63
Solubility 1.25e-02 mg/ml ; 2.36e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.84
Solubility 7.65e-06 mg/ml ; 1.44e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.14 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: hydroquinone
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.12
Molecule 175
SMILES COc1ccc(cc1)NCC(=O)Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C27H18N3O8S
Molecular weight 544.51 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 24
Fraction Csp3 0.11
Num. rotatable bonds 9
Num. H-bond acceptors 9
Num. H-bond donors 3
Molar Refractivity 141.91
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

203.06 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.86
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.88
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.27
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.48
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.98
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.50
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.63
Solubility 1.27e-01 mg/ml ; 2.33e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.73
Solubility 1.02e-02 mg/ml ; 1.87e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.53
Solubility 1.62e-06 mg/ml ; 2.97e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.00 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.22
Molecule 176
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C26H16N4O9S
Molecular weight 560.49 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.08
Num. rotatable bonds 9
Num. H-bond acceptors 10
Num. H-bond donors 4
Molar Refractivity 140.49
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

234.14 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.28
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.06
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.15
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

2.76
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.59
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.34
Solubility 2.55e-01 mg/ml ; 4.56e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.76
Solubility 9.77e-03 mg/ml ; 1.74e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.76
Solubility 9.74e-06 mg/ml ; 1.74e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.52 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 177
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C27H15N3O9S
Molecular weight 557.49 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 24
Fraction Csp3 0.07
Num. rotatable bonds 9
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 140.44
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

233.96 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

0.44
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.37
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-1.60
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.38
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.59
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.42
Solubility 2.10e-01 mg/ml ; 3.77e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-4.92
Solubility 6.69e-03 mg/ml ; 1.20e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.75
Solubility 9.80e-06 mg/ml ; 1.76e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.39 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.12
Molecule 178
SMILES CCOC(=O)c1ccc(cc1)NCC(=O)Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C29H20N3O9S
Molecular weight 586.55 g/mol
Num. heavy atoms 42
Num. arom. heavy atoms 24
Fraction Csp3 0.14
Num. rotatable bonds 11
Num. H-bond acceptors 10
Num. H-bond donors 3
Molar Refractivity 151.51
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

220.13 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.03
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

1.13
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-0.10
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.35
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

1.77
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.89
Solubility 7.64e-02 mg/ml ; 1.30e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.35
Solubility 2.64e-03 mg/ml ; 4.50e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.83
Solubility 8.74e-07 mg/ml ; 1.49e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.08 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 2 violations: Rotors>10, TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.39
Molecule 179
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1cccc2c1cccc2
Physicochemical Properties
Formula C30H18N3O7S
Molecular weight 564.54 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 28
Fraction Csp3 0.07
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 152.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

1.87
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.42
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.91
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.36
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.68
Solubility 1.18e-02 mg/ml ; 2.10e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.86
Solubility 7.74e-04 mg/ml ; 1.37e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.05
Solubility 5.04e-08 mg/ml ; 8.93e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.31
Molecule 180
SMILES O=C(Nc1scc(n1)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-])CNc1ccc2c(c1)cccc2
Physicochemical Properties
Formula C30H18N3O7S
Molecular weight 564.54 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 28
Fraction Csp3 0.07
Num. rotatable bonds 8
Num. H-bond acceptors 8
Num. H-bond donors 3
Molar Refractivity 152.92
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

193.83 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

2.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

2.16
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

1.42
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

0.44
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.91
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

2.39
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-4.68
Solubility 1.18e-02 mg/ml ; 2.10e-05 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-5.86
Solubility 7.74e-04 mg/ml ; 1.37e-06 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Moderately soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-10.05
Solubility 5.04e-08 mg/ml ; 8.93e-11 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Insoluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		Yes
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		Yes
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		Yes
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-8.21 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.31
Molecule 181
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccccc1
Physicochemical Properties
Formula C27H21N2O7
Molecular weight 485.46 g/mol
Num. heavy atoms 36
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 137.15
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-4.10
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.77
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.45
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.82
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.83
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.46
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.73
Solubility 9.05e+00 mg/ml ; 1.86e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.55
Solubility 1.36e+02 mg/ml ; 2.81e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.73
Solubility 9.00e-05 mg/ml ; 1.85e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.52 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.05
Molecule 182
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)C
Physicochemical Properties
Formula C28H23N2O7
Molecular weight 499.49 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.25
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 142.11
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.19
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.14
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.62
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.36
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.80
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.04
Solubility 4.57e+00 mg/ml ; 9.14e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.94
Solubility 5.79e+01 mg/ml ; 1.16e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.10
Solubility 3.93e-05 mg/ml ; 7.87e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.34 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 0 violation
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_E
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.17
Molecule 183
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)Cl
Physicochemical Properties
Formula C27H20ClN2O7
Molecular weight 519.91 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 142.16
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.72
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.14
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.79
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.35
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.47
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.71
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.33
Solubility 2.44e+00 mg/ml ; 4.69e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.21
Solubility 3.24e+01 mg/ml ; 6.23e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.31
Solubility 2.55e-05 mg/ml ; 4.91e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.28 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.04
Molecule 184
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)Br
Physicochemical Properties
Formula C27H20BrN2O7
Molecular weight 564.36 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 144.85
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.53
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.08
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.69
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.25
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.51
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.61
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.64
Solubility 1.29e+00 mg/ml ; 2.28e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.27
Solubility 3.04e+01 mg/ml ; 5.40e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.49
Solubility 1.83e-05 mg/ml ; 3.24e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.51 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.06
Molecule 185
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)C(Cl)(Cl)Cl
Physicochemical Properties
Formula C28H20Cl3N2O7
Molecular weight 602.83 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 19
Fraction Csp3 0.25
Num. rotatable bonds 7
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 156.38
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.05
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.15
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

0.27
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-2.77
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

5.59
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

0.18
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-3.37
Solubility 2.56e-01 mg/ml ; 4.24e-04 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-2.23
Solubility 3.52e+00 mg/ml ; 5.85e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.48
Solubility 1.97e-06 mg/ml ; 3.27e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.08 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: MW>600
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: alkyl_halide
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.17
Molecule 186
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)O
Physicochemical Properties
Formula C27H21N2O8
Molecular weight 501.46 g/mol
Num. heavy atoms 37
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 6
Num. H-bond acceptors 9
Num. H-bond donors 2
Molar Refractivity 139.17
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

155.35 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-3.07
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.12
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.74
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.32
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.35
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.58
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.60
Solubility 1.27e+01 mg/ml ; 2.53e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.61
Solubility 1.22e+02 mg/ml ; 2.43e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.14
Solubility 3.64e-04 mg/ml ; 7.26e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.86 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: XLOGP3<-2, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.09
Molecule 187
SMILES COc1ccc(cc1)N1CC[N+](CC1)CC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C28H23N2O8
Molecular weight 515.49 g/mol
Num. heavy atoms 38
Num. arom. heavy atoms 19
Fraction Csp3 0.25
Num. rotatable bonds 7
Num. H-bond acceptors 9
Num. H-bond donors 1
Molar Refractivity 143.64
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

144.35 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.26
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.80
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.44
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.12
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.91
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.14
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.81
Solubility 7.98e+00 mg/ml ; 1.55e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.71
Solubility 9.94e+01 mg/ml ; 1.93e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.83
Solubility 7.70e-05 mg/ml ; 1.49e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.72 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_C
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.18
Molecule 188
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)[N+](=O)O
Physicochemical Properties
Formula C27H21N3O9
Molecular weight 531.47 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 19
Fraction Csp3 0.22
Num. rotatable bonds 7
Num. H-bond acceptors 10
Num. H-bond donors 2
Molar Refractivity 142.22
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

175.43 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

0.00
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.40
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.65
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.81
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

1.68
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-1.23
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.52
Solubility 1.60e+01 mg/ml ; 3.01e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.74
Solubility 9.57e+01 mg/ml ; 1.80e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.06
Solubility 4.63e-04 mg/ml ; 8.71e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.25 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: XLOGP3<-2, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

1 alert: anil_di_alk_A
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

1 alert: oxygen-nitrogen_single_bond
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.17
Molecule 189
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc(cc1)C(=O)[O-]
Physicochemical Properties
Formula C28H20N2O9
Molecular weight 528.47 g/mol
Num. heavy atoms 39
Num. arom. heavy atoms 19
Fraction Csp3 0.21
Num. rotatable bonds 7
Num. H-bond acceptors 10
Num. H-bond donors 1
Molar Refractivity 142.16
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

175.25 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-10.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-2.24
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-3.08
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-4.14
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

3.31
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-3.34
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-1.60
Solubility 1.32e+01 mg/ml ; 2.49e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-0.91
Solubility 6.55e+01 mg/ml ; 1.24e-01 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-6.05
Solubility 4.66e-04 mg/ml ; 8.81e-07 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-11.11 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 2 violations: XLOGP3<-2, TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.11
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.12
Molecule 190
SMILES CCOC(=O)c1ccc(cc1)N1CC[N+](CC1)CC(=O)c1c2=c(c3c1c(=O)ccc3O)c(=O)c(cc2=O)CC(=O)[O-]
Physicochemical Properties
Formula C30H25N2O9
Molecular weight 557.53 g/mol
Num. heavy atoms 41
Num. arom. heavy atoms 19
Fraction Csp3 0.27
Num. rotatable bonds 9
Num. H-bond acceptors 10
Num. H-bond donors 1
Molar Refractivity 153.23
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

161.42 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.05
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-1.55
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-1.27
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.75
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.28
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.87
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.07
Solubility 4.76e+00 mg/ml ; 8.53e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.33
Solubility 2.59e+01 mg/ml ; 4.65e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-7.13
Solubility 4.15e-05 mg/ml ; 7.45e-08 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		Low
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		No
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-10.80 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		No; 2 violations: MW>500, NorO>10
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 3 violations: MW>480, WLOGP<-0.4, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		No; 1 violation: TPSA>140
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		No; 1 violation: TPSA>150
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.17
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 2 violations: MW>350, Rotors>7
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.39
Molecule 191
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1cccc2c1cccc2
Physicochemical Properties
Formula C31H23N2O7
Molecular weight 535.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 23
Fraction Csp3 0.19
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 154.65
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-2.56
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.52
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.29
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.21
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.85
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.35
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.86
Solubility 7.36e-01 mg/ml ; 1.37e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.85
Solubility 7.58e+00 mg/ml ; 1.42e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.35
Solubility 2.39e-06 mg/ml ; 4.47e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.94 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.28
Molecule 192
SMILES [O-]C(=O)Cc1cc(=O)c2=c(c1=O)c1c(O)ccc(=O)c1c2C(=O)C[N+]1CCN(CC1)c1ccc2c(c1)cccc2
Physicochemical Properties
Formula C31H23N2O7
Molecular weight 535.52 g/mol
Num. heavy atoms 40
Num. arom. heavy atoms 23
Fraction Csp3 0.19
Num. rotatable bonds 6
Num. H-bond acceptors 8
Num. H-bond donors 1
Molar Refractivity 154.65
TPSA

Topological Polar Surface Area: Calculated from
Ertl P. et al. 2000 J. Med. Chem.

135.12 Ų
Lipophilicity
Log Po/w (iLOGP)

iLOGP: in-house physics-based method implemented from
Daina A et al. 2014 J. Chem. Inf. Model.

-1.98
Log Po/w (XLOGP3)

XLOGP3: Atomistic and knowledge-based method calculated by
XLOGP program, version 3.2.2, courtesy of CCBG, Shanghai Institute of Organic Chemistry

-0.52
Log Po/w (WLOGP)

WLOGP: Atomistic method implemented from
Wildman SA and Crippen GM. 1999 J. Chem. Inf. Model.

-0.29
Log Po/w (MLOGP)

MLOGP: Topological method implemented from
Moriguchi I. et al. 1992 Chem. Pharm. Bull.
Moriguchi I. et al. 1994 Chem. Pharm. Bull.
Lipinski PA. et al. 2001 Adv. Drug. Deliv. Rev.

-3.21
Log Po/w (SILICOS-IT)

SILICOS-IT: Hybrid fragmental/topological method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

4.85
Consensus Log Po/w

Consensus Log Po/w: Average of all five predictions

-0.23
Water Solubility
Log S (ESOL)

ESOL: Topological method implemented from
Delaney JS. 2004 J. Chem. Inf. Model.

-2.86
Solubility 7.36e-01 mg/ml ; 1.37e-03 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Soluble
Log S (Ali)

Ali: Topological method implemented from
Ali J. et al. 2012 J. Chem. Inf. Model.

-1.85
Solubility 7.58e+00 mg/ml ; 1.42e-02 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Very soluble
Log S (SILICOS-IT)

SILICOS-IT: Fragmental method calculated by
FILTER-IT program, version 1.0.2, courtesy of SILICOS-IT, http://www.silicos-it.com

-8.35
Solubility 2.39e-06 mg/ml ; 4.47e-09 mol/l
Class

Solubility class: Log S scale
Insoluble < -10 < Poorly < -6 < Moderately < -4 < Soluble < -2 Very < 0 < Highly

Poorly soluble
Pharmacokinetics
GI absorption

Gatrointestinal absorption: according to the white of the BOILED-Egg

 		High
BBB permeant

BBB permeation: according to the yolk of the BOILED-Egg

 		No
P-gp substrate

P-glycoprotein substrate: SVM model built on 1033 molecules (training set)
and tested on 415 molecules (test set)
10-fold CV: ACC=0.72 / AUC=0.77
External: ACC=0.88 / AUC=0.94

 		Yes
CYP1A2 inhibitor

Cytochrome P450 1A2 inhibitor: SVM model built on 9145 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.83 / AUC=0.90
External: ACC=0.84 / AUC=0.91

 		No
CYP2C19 inhibitor

Cytochrome P450 2C19 inhibitor: SVM model built on 9272 molecules (training set)
and tested on 3000 molecules (test set)
10-fold CV: ACC=0.80 / AUC=0.86
External: ACC=0.80 / AUC=0.87

 		No
CYP2C9 inhibitor

Cytochrome P450 2C9 inhibitor: SVM model built on 5940 molecules (training set)
and tested on 2075 molecules (test set)
10-fold CV: ACC=0.78 / AUC=0.85
External: ACC=0.71 / AUC=0.81

 		No
CYP2D6 inhibitor

Cytochrome P450 2D6 inhibitor: SVM model built on 3664 molecules (training set)
and tested on 1068 molecules (test set)
10-fold CV: ACC=0.79 / AUC=0.85
External: ACC=0.81 / AUC=0.87

 		No
CYP3A4 inhibitor

Cytochrome P450 3A4 inhibitor: SVM model built on 7518 molecules (training set)
and tested on 2579 molecules (test set)
10-fold CV: ACC=0.77 / AUC=0.85
External: ACC=0.78 / AUC=0.86

 		No
Log Kp (skin permeation)

Skin permeation: QSPR model implemented from
Potts RO and Guy RH. 1992 Pharm. Res.

-9.94 cm/s
Druglikeness
Lipinski

Lipinski (Pfizer) filter: implemented from
Lipinski CA. et al. 2001 Adv. Drug Deliv. Rev.
MW ≤ 500
MLOGP ≤ 4.15
N or O ≤ 10
NH or OH ≤ 5

 		Yes; 1 violation: MW>500
Ghose

Ghose filter: implemented from
Ghose AK. et al. 1999 J. Comb. Chem.
160 ≤ MW ≤ 480
-0.4 ≤ WLOGP ≤ 5.6
40 ≤ MR ≤ 130
20 ≤ atoms ≤ 70

 		No; 2 violations: MW>480, MR>130
Veber

Veber (GSK) filter: implemented from
Veber DF. et al. 2002 J. Med. Chem.
Rotatable bonds ≤ 10
TPSA ≤ 140

 		Yes
Egan

Egan (Pharmacia) filter: implemented from
Egan WJ. et al. 2000 J. Med. Chem.
WLOGP ≤ 5.88
TPSA ≤ 131.6

 		No; 1 violation: TPSA>131.6
Muegge

Muegge (Bayer) filter: implemented from
Muegge I. et al. 2001 J. Med. Chem.
200 ≤ MW ≤ 600
-2 ≤ XLOGP ≤ 5
TPSA ≤ 150
Num. rings ≤ 7
Num. carbon > 4
Num. heteroatoms > 1
Num. rotatable bonds ≤ 15
H-bond acc. ≤ 10
H-bond don. ≤ 5

 		Yes
Bioavailability Score

Abbott Bioavailability Score: Probability of F > 10% in rat
implemented from
Martin YC. 2005 J. Med. Chem.

0.55
Medicinal Chemistry
PAINS

Pan Assay Interference Structures: implemented from
Baell JB. & Holloway GA. 2010 J. Med. Chem.

0 alert
Brenk

Structural Alert: implemented from
Brenk R. et al. 2008 ChemMedChem

0 alert
Leadlikeness

Leadlikeness: implemented from
Teague SJ. 1999 Angew. Chem. Int. Ed.
250 ≤ MW ≤ 350
XLOGP ≤ 3.5
Num. rotatable bonds ≤ 7

 		No; 1 violation: MW>350
Synthetic accessibility

Synthetic accessibility score: from 1 (very easy) to 10 (very difficult)
based on 1024 fragmental contributions (FP2) modulated by size and complexity penaties,
trained on 12’782’590 molecules and tested on 40 external molecules (r2 = 0.94)

 		4.28