Table 2 Comparison of the performance of ColdstartCPI and structure-based baselines on the DUD-E and LIT_PCBA benchmark
Datasets | Models | AUC (%) | BEDROC (%) | EF | ||
|---|---|---|---|---|---|---|
0.5% | 1% | 5% | ||||
DUD-E | Glide-SPa | 76.70 | 40.70 | 19.39 | 16.18 | 7.23 |
Vinab | 71.60 | —— | 9.13 | 7.32 | 4.44 | |
Sminac | 69.6 | —— | —— | —— | —— | |
NN-scorea | 68.30 | 12.20 | 4.16 | 4.02 | 3.12 | |
RFscorea | 65.21 | 12.41 | 4.90 | 4.52 | 2.98 | |
Pafnucya | 63.11 | 16.50 | 4.24 | 4.24 | 3.76 | |
OnionNeta | 59.71 | 8.62 | 2.84 | 2.84 | 2.20 | |
PLANETd | 71.60 | —— | 10.23 | 8.83 | 5.40 | |
ColdstartCPI | 76.53 | 20.78 | 21.61 | 15.15 | 6.42 | |
LIT_PCBA | Surflexe | 51.47 | —— | —— | 2.50 | —— |
Glide-SPd | 53.57 | 4.00 | 3.4 | 3.13 | 2.17 | |
PLANETd | 55.58 | —— | 3.64 | 3.28 | 2.01 | |
ConBAPf | 55.36 | —— | 3.89 | 2.53 | 1.65 | |
ColdstartCPI | 59.55 | 3.80 | 3.28 | 2.24 | 1.70 | |
