Table 1: Statistical performance of SVM classification models for substrate or inhibitor of pharmacokinetics-relevant protein, P-gp and CYP.

From: SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules

ModelSwissADMEPrevious models
TR/TS[a]ACCCV[b]AUCCV[c]ACCext[d]AUCext[e]TR/TSACCCVAUCCVACCextAUCextReference
P-gp substrate1033/4150.720.770.890.94544/n.c.0.71[f]n.c.n.c.n.c.53
      484/3000.64n.c.0.59n.c.71
      332/n.c.0.74[f]0.77[f]n.c.n.c.15
      212/1200.74n.c.0.88n.c.57
CYP1A2 inhibitor9145/30000.830.900.840.919145/3000n.c.n.c.0.880.9554
      12099/28040.82[f]n.c.0.680.8177
      7208/71280.88[g]n.c.n.c.0.9355
CYP2C19 inhibitor9272/30000.800.860.800.879272/3000n.c.n.c.0.850.9154
      11885/26910.79[f]n.c.0.810.8477
      6038/59230.81[g]n.c.n.c.0.8955
CYP2C9 inhibitor5940/20750.780.850.710.818720/3000n.c.n.c.0.830.9054
      12130/25790.78[f]n.c0.890.8677
      6627/65300.83[g]n.c.n.c.0.8955
CYP2D6 inhibitor3664/10680.790.850.810.879726/3000n.c.n.c.0.840.8854
      11881/28600.84[f]n.c.0.880.8877
      7788/77610.90[g]n.c.n.c.0.8555
CYP3A4 inhibitor7518/25790.770.850.780.868893/5135n.c.n.c.0.840.9254
      11536/70250.78[f]n.c.0.760.7877
      2334/67380.81[g]n.c.n.c.0.8755
  1. aNumber of molecules in the training set (TR) and in the test set (TS); b10-fold cross-validation accuracy; c10-fold cross-validation area under receiver operating characteristic (ROC) curve; dexternal validation accuracy; eexternal validation area under ROC curve; f5-fold cross-validation; g7-fold cross-validation.