Unconditional Protein Sequence Generation on Real Proteins Baseline
85.276pLDDTBioMatrix-1.7B
Evaluation results
| Method | Links | |
|---|---|---|
| BioMatrix-1.7BStructure prediction protocol=ESMFold, Model Scale=1.7B2026.06 | 85.276 | |
| BioMatrix-4BStructure prediction protocol=ESMFold, Model Scale=4B2026.06 | 85.012 | |
| DPLM2026.04 | 84 | |
| DPLM-Evo2026.04 | 83.6 | |
| DiMA2026.04 | 83.3 | |
| DPLM-650MStructure prediction protocol=ESMFold, Model Scale=650M2026.06 | 83.252 | |
| DPLM-2-650MStructure prediction protocol=ESMFold, Model Scale=650M2026.06 | 82.246 | |
| Dataset (real proteins)2026.04 | 80.7 | |
| nanoGPT2026.04 | 61 | |
| RITA2026.04 | 43.9 | |
| SeqDesign2026.04 | 43.1 | |
| DFM2026.04 | 37.8 | |
| EvoDiff-OADM2026.04 | 37.1 | |
| D3PM2026.04 | 36.7 | |
| EvoDiffStructure prediction protocol=ESMFold2026.06 | 35.846 | |
| Walk-Jump2026.04 | 32.4 | |
| proteinGAN2026.04 | 30.4 | |
| Random sequences2026.04 | 24.8 |