Protein language model rescue mutations highlight variant effects and structure in clinically relevant genes

TL;DR

This paper demonstrates that protein language models can identify rescue mutations that reveal structural features and variant effects in clinically relevant genes, surpassing traditional structure predictors.

Contribution

It introduces a systematic approach using protein language models to find compensatory mutations, uncovering structural insights missed by other methods.

Findings

Protein language models predict impactful rescue mutations.

Rescue mutations reveal new structural features.

Community efforts can improve clinical mutation predictions.

Abstract

Despite being self-supervised, protein language models have shown remarkable performance in fundamental biological tasks such as predicting impact of genetic variation on protein structure and function. The effectiveness of these models on diverse set of tasks suggests that they learn meaningful representations of fitness landscape that can be useful for downstream clinical applications. Here, we interrogate the use of these language models in characterizing known pathogenic mutations in curated, medically actionable genes through an exhaustive search of putative compensatory mutations on each variant's genetic background. Systematic analysis of the predicted effects of these compensatory mutations reveal unappreciated structural features of proteins that are missed by other structure predictors like AlphaFold. While deep mutational scan experiments provide an unbiased estimate of the mutational landscape, we encourage the community to generate and curate rescue mutation experiments to inform the design of more sophisticated co-masking strategies and leverage large language models more effectively for downstream clinical prediction tasks.