# Gene perturbation prediction on Adamson, Dixit, and Norman (test)

Source: [Gene perturbation prediction on Adamson, Dixit, and Norman (test)](<https://www.sota2.com/research/sota/gene-perturbation-prediction-on-adamson-dixit-and-norman-test>)

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

Data updated: 2026-02-26T05:34:40.891Z

Evaluation of models predicting gene perturbation outcomes using the test split derived from the Adamson, Dixit, and Norman study, assessed using metrics including PCCdelta.

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 | Gene perturbation prediction |
| Dataset (including any stated split) | Adamson, Dixit, and Norman (test) |
| Metric | Adamson Performance |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Adamson Performance](<https://www.sota2.com/research/sota/gene-perturbation-prediction-on-adamson-dixit-and-norman-test.md?metric=Adamson%20Performance>): Higher is better; unit: SCALAR.

- [Dixit Performance](<https://www.sota2.com/research/sota/gene-perturbation-prediction-on-adamson-dixit-and-norman-test.md?metric=Dixit%20Performance>): Higher is better; unit: SCALAR.

- [Norman Performance](<https://www.sota2.com/research/sota/gene-perturbation-prediction-on-adamson-dixit-and-norman-test.md?metric=Norman%20Performance>): Higher is better; unit: SCALAR.

- [Average PCCdelta](<https://www.sota2.com/research/sota/gene-perturbation-prediction-on-adamson-dixit-and-norman-test.md?metric=Average%20PCCdelta>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Adamson Performance | Result date | Report paper |
| --- | --- | --- | --- |
| scGPT + GRN (GRNFormer) | 0.622 | Not recorded | 68fa77381e209131cea71c19 |
| scGPT | 0.609 | Not recorded | 68fa77381e209131cea71c19 |
| scFoundation + GRN (GRNFormer) | 0.487 | Not recorded | 68fa77381e209131cea71c19 |
| scFoundation | 0.483 | Not recorded | 68fa77381e209131cea71c19 |

### scGPT + GRN (GRNFormer)

Adamson Performance: 0.622. Result ID: 69946dd73a6c3148491d2c96.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

Recorded evaluation configuration: {"Backbone":"scGPT","GRN Enhancement":true,"Fine-tuning Strategy":"full parameter updates","Pre-training Protocol":"SEA-AD multiome data"}. This record does not assert independent reproduction.

### scGPT

Adamson Performance: 0.609. Result ID: 69946dd73a6c3148491d2c95.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

Recorded evaluation configuration: {"Backbone":"scGPT","GRN Enhancement":false,"Fine-tuning Strategy":"full parameter updates","Pre-training Protocol":"SEA-AD multiome data"}. This record does not assert independent reproduction.

### scFoundation + GRN (GRNFormer)

Adamson Performance: 0.487. Result ID: 69946dd73a6c3148491d2c98.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

Recorded evaluation configuration: {"Backbone":"scFoundation","GRN Enhancement":true,"Fine-tuning Strategy":"parameter freezing","Pre-training Protocol":"continued pretraining from public checkpoint"}. This record does not assert independent reproduction.

### scFoundation

Adamson Performance: 0.483. Result ID: 69946dd73a6c3148491d2c97.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77381e209131cea71c19.md>)

Recorded evaluation configuration: {"Backbone":"scFoundation","GRN Enhancement":false,"Fine-tuning Strategy":"parameter freezing","Pre-training Protocol":"continued pretraining from public checkpoint"}. This record does not assert independent reproduction.

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