Network Analysis and Ligand Based Pharmacophore Modeling to Discover Small Molecule against Glioblastoma Multieforme - Supporting Information.docx

  

Fig.S1. Gene ontology analysis for cluster-Ⅰ genes. The nodes are represented as colored and uncolored, where the colored node acts as an overrepresented function while the uncolored node acts as a parent node.

 Fig.S2. KEGG pathway analysis of C-SRC. VEGF signaling pathway (hsa04370), EGFR inhibitor resistance pathway (hsa01521), ERBB signaling pathway (hsa04012) and GABAergic synapse pathway (hsa04727) contain C-SRC as a significant mediator in GBM regulation.  

  

Table.S1. Topology analysis of cluster-Ⅰ genes using cytoHubba. SRC gene posses highest degree value 10 and betweenness value 23.833 representing the most important gene among other in cluster-Ⅰ.

  

Table.S2. Protein data bank structures allocated with C-SRC gene. PDB ID 2SRC and 1FMK was found to have better characteristics while 2SRC have an inhibitor in its active site. 

  

Fig.S3. Ramachandran plot of C-SRC (PDB ID: 2SRC). Almost every amino acids were present in allowed region (colored area) representing the protein structure have good torsional stability. 

  

Table.S3. Docking score of top 10 drugs with C-SRC (PDB ID: 2SRC). Pemetrexed has obtained the best binding affinity towards the receptor in docking with Glide (Schrodinger) and CDocker (Discovery Studio).

  

Fig.S4. The three dimensional interaction view of Pemetrexed in binding site. Green colored dotted lines indicate the drug is having good clashes with the receptor and does not have any unwanted steric interactions.

  

Table.S4. The estimated activity of training set molecules in µM level. Error-values represent the ratio of experimental activity against the estimated activity. Negative error values indicate the estimated activity is lower than experimental activity and positive error values indicate estimated activity is higher than experimental activity. Fit values indicate the mapping of the training set molecules in the predicted pharmacophore. The activity scale classifies molecules into three sets based on the experimental activity. Highly active (+++), moderately active (++), low active (+).

 Fig.S5. Correlation between experimental and estimated activity for test set molecules. A q2 value of 0.8548 obtained shows training and test set molecules posse’s better correlation between them 

  

Fig.S6. Overlay of best-fit ligand after screening all ligands (A) and Pemetrexed (B) in Hypo1 model

  

Table.S5. Docking score and MM-GBSA score of best molecules obtained after virtual screening. 

  

Fig.S7. Receptor pocket surface of protein tyrosine kinase (PDB: 2SRC). Each color represents the binding property of active site. Aliphatic (green color), hydrogen donor (blue color), hydrogen acceptor (red color), and aromatic region (white color) were present in the active site.

  

Fig.S8. Interaction of NSC-13377 in the binding site of protein tyrosine kinase C-SRC. Three-dimensional interaction (B) reveals the ligand is not having any steric clash and two-dimensional interactions (A) depict the type of interaction involved.

  

Fig.S9. Best molecules obtained after screening and docking. Binding free energy for the molecules was calculated using the MM-GBSA method is depicted below the structures.

  

Fig.S10. Enrichment factor curve for active ligands

  

Table.S6. Enrichment calculation of virtual screening protocol. ROC value of 0.95 revealed the protocol is validated

  

Table.S7. ADMET and toxicity level comparison of best hit ligands with Pemetrexed. NSC-13377 and NSC-400902 show better solubility and high BBB penetration than Pemetrexed

  

Fig.S11. ADMET plot comparison of best-fit molecules against Pemetrexed. The eclipse represents 95% and 99% confidence limit of human intestinal absorption and blood-brain barrier (BBB) permeation.