Skip to main content
. 2023 Sep 29;63(20):6396–6411. doi: 10.1021/acs.jcim.3c00939

Table 1. Theoretical Evaluation of LS and DS Generated Modelsa.

Model Actives Decoy TPs FPs TNs FNs Accuracy Sensitivity Specificity YoA EF
SB1 7 49 7 49 4852 88 0.97 0.07 0.99 0.13 6.57
SB2 6 69 6 69 4832 89 0.97 0.06 0.99 0.08 4.21
LB3 12 51 12 51 4850 83 0.97 0.13 0.99 0.19 10.02
LB4 8 15 8 15 4886 87 0.98 0.08 1.00 0.35 18.29
LB5 13 16 13 16 4885 82 0.98 0.14 1.00 0.45 23.57
LB6 5 24 5 24 4877 90 0.98 0.05 1.00 0.17 9.07
LB7 4 32 4 32 4869 91 0.98 0.04 0.99 0.11 5.84
LB8 3 18 3 18 4883 92 0.98 0.03 1.00 0.14 7.51
LB9 5 52 5 52 4849 90 0.97 0.05 0.99 0.09 4.61
LB10 5 58 5 58 4843 90 0.97 0.05 0.99 0.08 4.17
LB11 9 45 9 45 4856 86 0.97 0.09 0.99 0.17 8.76
DS2 10 49 10 49 4852 85 0.97 0.11 0.99 0.17 8.91
DS5 11 57 11 57 4844 84 0.97 0.12 0.99 0.16 8.51
DS8 10 74 10 74 4827 85 0.97 0.11 0.98 0.12 6.26
DS9 9 64 9 64 4837 86 0.97 0.09 0.99 0.12 6.48
DS12 15 86 15 86 4815 80 0.97 0.16 0.98 0.15 7.81
Combined 89 677 89 677 4224 6 0.86 0.94 0.86 0.12 6.11
a

Each model was evaluated in relation to EF, YoA, and number of TPs.