DBPP-Predictor: a novel strategy for prediction of chemical drug-likeness based on property profiles

Table 2 Ten-fold cross-validation results for models based on different representation

Representation	Accuracy	Precision	Recall	AUC	SP
Descriptors	0.968 ± 0.001	0.965 ± 0.003	0.972 ± 0.001	0.994 ± 0.000	0.964 ± 0.003
FP	0.972 ± 0.002	0.976 ± 0.002	0.967 ± 0.002	0.994 ± 0.000	0.976 ± 0.002
GCN	0.901 ± 0.045	0.849 ± 0.074	0.984 ± 0.016	0.988 ± 0.002	0.818 ± 0.104
QED	0.627 ± 0.004	0.637 ± 0.005	0.591 ± 0.001	0.684 ± 0.002	0.663 ± 0.007
ADMET Property	0.984 ± 0.000	0.992 ± 0.000	0.975 ± 0.000	0.996 ± 0.000	0.992 ± 0.000
Property Profiles	0.903 ± 0.001	0.905 ± 0.002	0.899 ± 0.003	0.961 ± 0.001	0.906 ± 0.002

ISSN: 1758-2946