Characterizing the combined effects of cytochrome P450 missense variation within star allele definitions
Supplementary Figure 1. Comparison of Fold-X computed (∆∆G) and ENCoM-computed (∆∆Svib). Each haplotype is represented by a maroon square and their defining missense variants are represented by circles. Figure points that fall within the highlighted region are expected to have neutral effects on protein stability . ∆∆G predicts most of the variants and haplotypes as destabilising to the protein encoded protein structures. Both ∆∆G and ∆∆Svib seem to exhibit a destabilising bias for both haplotype and variants. However, ∆∆Svib predicted more haplotypes and variants to have neutral effects on protein stability than ∆∆G (see also Figure 3A).
Supplementary Table 1. This table presents all 261 missense variants analysed in this study, together with their FoldX and ENCoM-derived energy scores. A combined ∆∆G (combination of ENCoM and FoldX) is presented, from which a prediction of the effect of each variant is made based on the threshold (-0.5 < combined ∆∆G < 0.5) . Predictions from commonly used Ensembl Variant Effect Predictor (VEP; https://www.ensembl.org/ToolsVEP)  plugins as well as a consensus (2 from 3 rule) has been included for comparison with combined ∆∆G predictions
Supplementary Table 2. CYP star allele/haplotype NMA analysis. This table presents 92 CYP star alleles together with the defining missense variants for each star allele. Moreover, a CPIC functional classification is included together with the NMA-based combined ∆∆G analysis.
Supplementary Table 3. SWAAT analysis of CYP missense variant structural consequence.