Blind self-docking on PDBbind v2020 (test)
1.2Ligand RMSD (25%)FABind+
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| FABind+Docking approach=Regression2024.03 | 1.2 | 2.6 | 5.8 | 5.2 | 0.435 | 0.711 | 0.4 | 1 | 2.9 | 3.5 | 0.675 | 0.84 | 0.16 | |
| FABind+Docking approach=Sampling, Sample size=402024.03 | 1.2 | 2.4 | 5.6 | 5.2 | 0.449 | 0.713 | 0.5 | 1 | 2.7 | 3.5 | 0.683 | 0.852 | 6.4 | |
| FABind+Docking approach=Sampling, Sample size=102024.03 | 1.3 | 2.7 | 5.4 | 5.2 | 0.424 | 0.716 | 0.5 | 1.1 | 2.8 | 3.5 | 0.678 | 0.846 | 1.6 | |
| DIFFDOCKDocking approach=Deep learning-based method, Sample size=402024.03 | 1.4 | 3.3 | 7.3 | — | 0.382 | 0.632 | 0.5 | 1.2 | 3.2 | — | 0.645 | 0.805 | 82.83 | |
| DIFFDOCKDocking approach=Deep learning-based method, Sample size=102024.03 | 1.5 | 3.6 | 7.1 | — | 0.35 | 0.617 | 0.5 | 1.2 | 3.3 | — | 0.631 | 0.807 | 20.81 | |
| FABINDDocking approach=Deep learning-based method2024.03 | 1.7 | 3.1 | 6.7 | 6.4 | 0.331 | 0.642 | 0.7 | 1.3 | 3.6 | 4.7 | 0.603 | 0.802 | 0.12 | |
| E3BINDDocking approach=Deep learning-based method2024.03 | 2.1 | 3.8 | 7.8 | 7.2 | 0.234 | 0.6 | 0.8 | 1.5 | 4 | 5.1 | 0.6 | 0.788 | 0.44 | |
| QVINA-WDocking approach=Traditional docking software, CPU-only=true2024.03 | 2.5 | 7.7 | 23.7 | 13.6 | 0.209 | 0.402 | 0.9 | 3.7 | 22.9 | 11.9 | 0.41 | 0.546 | 49 | |
| GLIDEDocking approach=Traditional docking software, CPU-only=true2024.03 | 2.6 | 9.3 | 28.1 | 16.2 | 0.218 | 0.336 | 0.8 | 5.6 | 26.9 | 14.4 | 0.361 | 0.487 | 1,405 | |
| TANKBINDDocking approach=Deep learning-based method2024.03 | 2.6 | 4.2 | 7.6 | 7.8 | 0.176 | 0.578 | 0.8 | 1.7 | 4.3 | 5.9 | 0.55 | 0.778 | 0.87 | |
| GNINADocking approach=Traditional docking software2024.03 | 2.8 | 8.7 | 22.1 | 13.3 | 0.212 | 0.371 | 1 | 4.5 | 21.2 | 11.5 | 0.36 | 0.52 | 146 | |
| SMINADocking approach=Traditional docking software, CPU-only=true2024.03 | 3.8 | 8.1 | 17.9 | 12.1 | 0.135 | 0.339 | 1.3 | 3.7 | 16.2 | 9.8 | 0.38 | 0.559 | 146 | |
| EQUIBINDDocking approach=Deep learning-based method2024.03 | 3.8 | 6.2 | 10.3 | 8.2 | 0.055 | 0.391 | 1.3 | 2.6 | 7.4 | 5.6 | 0.4 | 0.675 | 0.03 | |
| VINADocking approach=Traditional docking software, CPU-only=true2024.03 | 5.7 | 10.7 | 21.4 | 14.7 | 0.055 | 0.212 | 1.9 | 6.2 | 20.1 | 12.1 | 0.265 | 0.471 | 205 |