Protein-Ligand Affinity Prediction 벤치마크
Protein-Ligand Affinity Prediction on PDBbind
RMSE 낮을수록 좋음
- 2021-07-21 — SIGN: RMSE 1.316
- 2023-11-19 — PAMNet: RMSE 1.263
- 2024-02-08 — PLAPT: RMSE 1.211
- 2024-11-06 — BAPULM: RMSE 0.898
| Rank | Model | RMSE | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | BAPULM | 0.898±0.0172 | BAPULM: Binding Affinity Prediction using Language Models | radh55sh/BAPULM | 2024 |
| 2 | PLAPT | 1.211 | PLAPT: Protein-Ligand Binding Affinity Prediction Using Pretrained Transformers | trrt-good/WELP-PLAPT | 2024 |
| 3 | PAMNet | 1.263 | A Universal Framework for Accurate and Efficient Geometric Deep Learning of Molecular Systems | XieResearchGroup/Physics-aware-Multiplex-GNN | 2023 |
| 4 | LightGBM | 1.316 | High Performance of Gradient Boosting in Binding Affinity Prediction | 2022 | |
| 4 | SIGN | 1.316 | Structure-aware Interactive Graph Neural Networks for the Prediction of Protein-Ligand Binding Affinity | agave233/SIGN | 2021 |
| 6 | DimeNet | 1.453 | Structure-aware Interactive Graph Neural Networks for the Prediction of Protein-Ligand Binding Affinity | agave233/SIGN | 2021 |
| 7 | GraphDTA | 1.562 | Structure-aware Interactive Graph Neural Networks for the Prediction of Protein-Ligand Binding Affinity | agave233/SIGN | 2021 |