Table 5.

Performances with LRs using DeepSnap_python system of the three MIE targets.

PubChem		720725_GR_Ant	1347030_TRHR	1347032_TGF_Beta
Assay AID		Train:Valid:Test = 7:1:2	Train:Valid:Test = 3:1:2	Train:Valid:Test = 7:1:2
ROC_AUC	average	0.983 ± 0.032	0.929 ± 0.003	0.918 ± 0.005
	max_ROC_AUC	0.983	0.934	0.925
	max_BS	125	14	28
BAC	average	0.866 ± 0.033	0.877 ± 0.004	0.848 ± 0.007
	max_BAC	0.930	0.881	0.862
	max_BS	125	22	44
MCC	average	0.444 ± 0.056	0.194 ± 0.004	0.368 ± 0.011
	max_MCC	0.604	0.200	0.390
	max_BS	200	14	28
Acc	average	0.908 ± 0.021	0.855 ± 0.005	0.810 ± 0.011
	max_Acc	0.954	0.863	0.835
	max_BS	200	14	20
loss_train	average	0.045 ± 0.033	0.322 ± 0.013	0.097 ± 0.047
	min_loss	0.019	0.301	0.037
	min_BS	48	14	20
loss_test	average	0.119 ± 0.025	0.314 ± 0.022	0.203 ± 0.023
	min_loss	0.073	0.255	0.172
	min_BS	120	2	34
PR_AUC	average	0.654 ± 0.087	0.136 ± 0.011	0.431 ± 0.032
	max_PR_AUC	0.800	0.154	0.476
	max_BS	290	14	28
F	average	0.930 ± 0.014	0.914 ± 0.003	0.860 ± 0.008
	max_F	0.961	0.919	0.877
	max_BS	200	14	20