# Mutation Fitness Prediction on ProteinGym substitution benchmark (217 proteins) (fold 4 test)

Source: [Mutation Fitness Prediction on ProteinGym substitution benchmark (217 proteins) (fold 4 test)](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-15T07:52:18.237Z

Evaluation of model performance on predicting mutation fitness using the specific 217-protein substitution benchmark subset of ProteinGym (fold 4 test), evaluated using classification and ranking metrics (AUROC, AUPRC, ACC, mAP).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Mutation Fitness Prediction |
| Dataset (including any stated split) | ProteinGym substitution benchmark (217 proteins) (fold 4 test) |
| Metric | AUROC |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [AUROC](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=AUROC>): Higher is better; unit: PERCENT.

- [AUPRC](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=AUPRC>): Higher is better; unit: PERCENT.

- [Accuracy](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [mF1](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=mF1>): Higher is better; unit: PERCENT.

- [mAR](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=mAR>): Higher is better; unit: PERCENT.

- [mAP](<https://www.sota2.com/research/sota/mutation-fitness-prediction-on-proteingym-substitution-benchmark-217-proteins-fold-4-test.md?metric=mAP>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AUROC | Result date | Report paper |
| --- | --- | --- | --- |
| TriFit | 89.74 | Not recorded | 69df1b9c7d0172140ec7b967 |
| Kermut | 86.44 | Not recorded | 69df1b9c7d0172140ec7b967 |
| ProteinNPT | 84.39 | Not recorded | 69df1b9c7d0172140ec7b967 |
| Emb + MSA-T | 83.34 | Not recorded | 69df1b9c7d0172140ec7b967 |
| Emb + ESM-1v | 80.62 | Not recorded | 69df1b9c7d0172140ec7b967 |
| OHE + TranceptEVE | 77.44 | Not recorded | 69df1b9c7d0172140ec7b967 |
| OHE + MSA-T | 76.44 | Not recorded | 69df1b9c7d0172140ec7b967 |
| OHE + ESM-1v | 75.81 | Not recorded | 69df1b9c7d0172140ec7b967 |
| OHE (no aug.) | 63.87 | Not recorded | 69df1b9c7d0172140ec7b967 |

### TriFit

AUROC: 89.74. Result ID: 69df426447f34ce924c7c3e1.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Model":"TriFit"}. This record does not assert independent reproduction.

### Kermut

AUROC: 86.44. Result ID: 69df426447f34ce924c7c3e0.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Model":"Kermut"}. This record does not assert independent reproduction.

### ProteinNPT

AUROC: 84.39. Result ID: 69df426447f34ce924c7c3df.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Model":"ProteinNPT"}. This record does not assert independent reproduction.

### Emb + MSA-T

AUROC: 83.34. Result ID: 69df426447f34ce924c7c3de.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Embedding":"MSA Transformer"}. This record does not assert independent reproduction.

### Emb + ESM-1v

AUROC: 80.62. Result ID: 69df426447f34ce924c7c3dd.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Embedding":"ESM-1v"}. This record does not assert independent reproduction.

### OHE + TranceptEVE

AUROC: 77.44. Result ID: 69df426447f34ce924c7c3dc.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Augmentation":"TranceptEVE"}. This record does not assert independent reproduction.

### OHE + MSA-T

AUROC: 76.44. Result ID: 69df426447f34ce924c7c3db.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Augmentation":"MSA Transformer"}. This record does not assert independent reproduction.

### OHE + ESM-1v

AUROC: 75.81. Result ID: 69df426447f34ce924c7c3da.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Augmentation":"ESM-1v"}. This record does not assert independent reproduction.

### OHE (no aug.)

AUROC: 63.87. Result ID: 69df426447f34ce924c7c3d9.

[Reporting paper](<https://www.sota2.com/research/paper/69df1b9c7d0172140ec7b967.md>)

Recorded evaluation configuration: {"Augmentation":"None"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.