Performance of the 9 Bayesian models for the training set and test set using different combinations of molecular descriptors.

Model	Descriptors	Number of descriptors	Training set				Test set
			SE	SP	Q	MCC	SE	SP	Q	MCC
NB-a	DS_2D_MP	16	0.872	0.866	0.868	0.686	0.853	0.826	0.833	0.619
NB-b	MP + ECFP_4	17	0.985	0.992	0.991	0.975	0.982	0.984	0.983	0.956
NB-c	MP + ECFP_6	17	0.988	0.997	0.995	0.985	0.982	0.989	0.987	0.966
NB-d	MP + EPFP_4	17	0.937	0.932	0.933	0.833	0.937	0.904	0.912	0.789
NB-e	MP + EPFP_6	17	0.962	0.956	0.957	0.891	0.960	0.918	0.928	0.828
NB-f	MP + FCFP_4	17	0.979	0.978	0.978	0.944	0.961	0.962	0.961	0.901
NB-g	MP + FCFP_6	17	0.987	0.990	0.990	0.973	0.971	0.987	0.983	0.955
NB-h	MP + FPFP-4	17	0.952	0.948	0.949	0.870	0.938	0.920	0.924	0.815
NB-i	MP + FPFP-6	17	0.955	0.973	0.968	0.917	0.958	0.931	0.938	0.847