Exploring Large Protein Language Models in Constrained Evaluation Scenarios within the FLIP Benchmark
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913671127498752 |
|---|---|
| author | Mollon, Manuel F. Gonzalez-Rodriguez, Joaquin Lozano-Diez, Alicia Ramos, Daniel Toledano, Doroteo T. |
| author_facet | Mollon, Manuel F. Gonzalez-Rodriguez, Joaquin Lozano-Diez, Alicia Ramos, Daniel Toledano, Doroteo T. |
| contents | In this study, we expand upon the FLIP benchmark-designed for evaluating protein fitness prediction models in small, specialized prediction tasks-by assessing the performance of state-of-the-art large protein language models, including ESM-2 and SaProt on the FLIP dataset. Unlike larger, more diverse benchmarks such as ProteinGym, which cover a broad spectrum of tasks, FLIP focuses on constrained settings where data availability is limited. This makes it an ideal framework to evaluate model performance in scenarios with scarce task-specific data. We investigate whether recent advances in protein language models lead to significant improvements in such settings. Our findings provide valuable insights into the performance of large-scale models in specialized protein prediction tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_18223 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Exploring Large Protein Language Models in Constrained Evaluation Scenarios within the FLIP Benchmark Mollon, Manuel F. Gonzalez-Rodriguez, Joaquin Lozano-Diez, Alicia Ramos, Daniel Toledano, Doroteo T. Machine Learning Artificial Intelligence In this study, we expand upon the FLIP benchmark-designed for evaluating protein fitness prediction models in small, specialized prediction tasks-by assessing the performance of state-of-the-art large protein language models, including ESM-2 and SaProt on the FLIP dataset. Unlike larger, more diverse benchmarks such as ProteinGym, which cover a broad spectrum of tasks, FLIP focuses on constrained settings where data availability is limited. This makes it an ideal framework to evaluate model performance in scenarios with scarce task-specific data. We investigate whether recent advances in protein language models lead to significant improvements in such settings. Our findings provide valuable insights into the performance of large-scale models in specialized protein prediction tasks. |
| title | Exploring Large Protein Language Models in Constrained Evaluation Scenarios within the FLIP Benchmark |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2501.18223 |