Inverse Folding on OpenFold n=50 (test)
67TM-ScoreProteinMPNN + DPO (scaled)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ProteinMPNN + DPO (scaled)Reward scaling=true2024.10 | 67 | 28 | 31 | |
| ProteinMPNN + DPO (α = 0.1)Diversity penalty (alpha)=0.12024.10 | 67 | 32 | 31 | |
| ProteinMPNN + DPO (scaled, α = 0.1)Reward scaling=true, Diversity penalty (alpha)=0.12024.10 | 67 | 31 | 32 | |
| ProteinMPNN + DPOReward scaling=false, Diversity penalty (alpha)=0.02024.10 | 66 | 27 | 33 | |
| GVP-GNN2024.10 | 62 | 19 | 27 | |
| ProteinMPNN2024.10 | 62 | 31 | 23 | |
| ESM-IF12024.10 | 61 | 0 | 31 |