Layer Probing Improves Kinase Functional Prediction with Protein Language Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911294225907712 |
|---|---|
| author | Kumar, Ajit Jha, IndraPrakash |
| author_facet | Kumar, Ajit Jha, IndraPrakash |
| contents | Protein language models (PLMs) have transformed sequence-based protein analysis, yet most applications rely only on final-layer embeddings, which may overlook biologically meaningful information encoded in earlier layers. We systematically evaluate all 33 layers of ESM-2 for kinase functional prediction using both unsupervised clustering and supervised classification. We show that mid-to-late transformer layers (layers 20-33) outperform the final layer by 32 percent in unsupervised Adjusted Rand Index and improve homology-aware supervised accuracy to 75.7 percent. Domain-level extraction, calibrated probability estimates, and a reproducible benchmarking pipeline further strengthen reliability. Our results demonstrate that transformer depth contains functionally distinct biological signals and that principled layer selection significantly improves kinase function prediction. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_00376 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Layer Probing Improves Kinase Functional Prediction with Protein Language Models Kumar, Ajit Jha, IndraPrakash Quantitative Methods Artificial Intelligence Machine Learning 92C40, 68T50 I.2.6; I.5.1; J.3 Protein language models (PLMs) have transformed sequence-based protein analysis, yet most applications rely only on final-layer embeddings, which may overlook biologically meaningful information encoded in earlier layers. We systematically evaluate all 33 layers of ESM-2 for kinase functional prediction using both unsupervised clustering and supervised classification. We show that mid-to-late transformer layers (layers 20-33) outperform the final layer by 32 percent in unsupervised Adjusted Rand Index and improve homology-aware supervised accuracy to 75.7 percent. Domain-level extraction, calibrated probability estimates, and a reproducible benchmarking pipeline further strengthen reliability. Our results demonstrate that transformer depth contains functionally distinct biological signals and that principled layer selection significantly improves kinase function prediction. |
| title | Layer Probing Improves Kinase Functional Prediction with Protein Language Models |
| topic | Quantitative Methods Artificial Intelligence Machine Learning 92C40, 68T50 I.2.6; I.5.1; J.3 |
| url | https://arxiv.org/abs/2512.00376 |