Skip to main content

Advertisement

Table 3 Results of the QSAR training and the virtual screening without consideration of the applicability domain

From: Estimation of the applicability domain of kernel-based machine learning models for virtual screening

  Thrombin Factor Xa PDGFRβ
  OAK FlexOAK MARG OAK FlexOAK MARG OAK FlexOAK MARG
  QSAR Training
Q2 0.49 ± 0.17 0.52 ± 0.15 0.52 ± 0.11 0.82 ± 0.08 0.85 ± 0.06 0.78 ± 0.1 0.38 ± 0.24 0.37 ± 0.29 0.40 ± 0.22
RMSE 0.55 ± 0.22 0.51 ± 0.15 0.50 ± 0.17 0.63 ± 0.24 0.53 ± 0.17 0.77 ± 0.29 0.34 ± 0.25 0.38 ± 0.30 0.32 ± 0.22
  Virtual Screening without Applicability Domain Estimation
AUC 0.45 0.51 0.54 0.74 0.68 0.54 0.58 0.54 0.25
BEDROC (100.0) 0.42 0.29 0.29 0.32 0.28 0.00 0.13 0.10 0.07
BEDROC (53.6) 0.37 0.29 0.27 0.31 0.25 0.01 0.12 0.09 0.06
BEDROC (32.2) 0.37 0.29 0.27 0.33 0.24 0.01 0.12 0.09 0.06
BEDROC (20.0) 0.35 0.30 0.28 0.37 0.26 0.02 0.13 0.11 0.05