Table 4.

Performance comparison. ROC-AUC, PR-AUC, Sensitivity, and Specificity were measured over five random runs to accurately compare the performance with that of the cited MolTrans paper. CI and $r_m^2$ compare the performance of the proposed model with the performance when a metric is added to the published MolTrans model. The highest performances among the metrics used to evaluate each dataset are shown in bold on a gray background.

Method	ROC-AUC	PR-AUC	Sensitivity	Specificity	CI	${\mathit{r}}_{\mathit{m}}^2$
Dataset 1. BIOSNAP
MolTrans	0.895 ± 0.002	0.901 ± 0.004	0.775 ± 0.032	0.851 ± 0.014	0.889	0.449
NP-Model-BS	0.882 ± 0.004	0.871 ± 0.015	0.779 ± 0.020	0.850 ± 0.012	0.895	0.428
CP-Model-BS	0.881 ± 0.009	0.859 ± 0.017	0.811 ± 0.018	0.835 ± 0.008	0.891	0.406
PP-Model-BS	0.893 ± 0.003	0.874 ± 0.006	0.803 ± 0.033	0.851 ± 0.019	0.896	0.425
FP-Model-BS	0.914 ± 0.006	0.900 ± 0.007	0.862 ± 0.025	0.847 ± 0.007	0.913	0.467
NP-Model-INT	0.877 ± 0.007	0.860 ± 0.010	0.785 ± 0.007	0.842 ± 0.008	0.897	0.421
CP-Model-INT	0.875 ± 0.006	0.851 ± 0.009	0.775 ± 0.023	0.844 ± 0.016	0.885	0.401
PP-Model-INT	0.895 ± 0.003	0.880 ± 0.008	0.802 ± 0.018	0.852 ± 0.009	0.896	0.435
FP-Model-INT	0.910 ± 0.012	0.897 ± 0.014	0.830 ± 0.029	0.863 ± 0.011	0.911	0.447
Dataset 2. DAVIS
MolTrans	0.907 ± 0.002	0.404 ± 0.016	0.800 ± 0.022	0.876 ± 0.013	0.903	0.156
NP-Model-DV	0.870 ± 0.003	0.283 ± 0.005	0.738 ± 0.030	0.871 ± 0.026	0.875	0.118
CP-Model-DV	0.882 ± 0.006	0.250 ± 0.023	0.744 ± 0.021	0.888 ± 0.019	0.878	0.117
PP-Model-DV	0.866 ± 0.003	0.263 ± 0.007	0.747 ± 0.020	0.856 ± 0.012	0.864	0.115
FP-Model-DV	0.920 ± 0.002	0.395 ± 0.007	0.824 ± 0.026	0.889 ± 0.015	0.917	0.167
NP-Model-INT	0.899 ± 0.008	0.322 ± 0.030	0.814 ± 0.039	0.857 ± 0.028	0.892	0.141
CP-Model-INT	0.904 ± 0.011	0.351 ± 0.035	0.814 ± 0.030	0.859 ± 0.020	0.917	0.169
PP-Model-INT	0.923 ± 0.005	0.417 ± 0.028	0.844 ± 0.017	0.876 ± 0.021	0.916	0.162
FP-Model-INT	0.942 ± 0.005	0.517 ± 0.017	0.903 ± 0.017	0.866 ± 0.015	0.940	0.201
Dataset 3. BindingDB
MolTrans	0.914 ± 0.001	0.622 ± 0.007	0.797 ± 0.005	0.896 ± 0.007	0.899	0.267
NP-Model-BDB	0.891 ± 0.005	0.515 ± 0.014	0.774 ± 0.012	0.897 ± 0.013	0.899	0.309
CP-Model-BDB	0.914 ± 0.003	0.585 ± 0.021	0.803 ± 0.011	0.904 ± 0.010	0.907	0.320
PP-Model-BDB	0.897 ± 0.003	0.557 ± 0.013	0.775 ± 0.019	0.900 ± 0.009	0.913	0.324
FP-Model-BDB	0.922 ± 0.001	0.623 ± 0.010	0.814 ± 0.025	0.916 ± 0.016	0.927	0.365
NP-Model-INT	0.904 ± 0.001	0.574 ± 0.008	0.766 ± 0.015	0.910 ± 0.015	0.907	0.315
CP-Model-INT	0.909 ± 0.005	0.600 ± 0.019	0.787 ± 0.008	0.907 ± 0.008	0.918	0.330
PP-Model-INT	0.918 ± 0.001	0.607 ± 0.012	0.787 ± 0.014	0.920 ± 0.010	0.916	0.344
FP-Model-INT	0.926 ± 0.001	0.639 ± 0.018	0.802 ± 0.022	0.928 ± 0.013	0.926	0.362