Virtual Screening on DUD-E 10 diverse targets
99.1Mean AUC1-NN Tanimoto
Evaluation results
| Method | Links | ||
|---|---|---|---|
| 1-NN TanimotoType=No learning, Exploits structural bias=true2026.04 | 99.1 | — | |
| Per-target RFType=Fingerprint, Trained per-target on ChEMBL=true2026.04 | 87.5 | 80 | |
| NESTDRUGType=GNN+FiLM2026.04 | 85 | 20 | |
| 3D-CNNType=3D CNN2026.04 | 83 | — | |
| GNN-VSType=GNN2026.04 | 82.5 | — | |
| AtomNetType=3D CNN2026.04 | 81.8 | — |