Ranking protein-peptide binding affinities with protein language models
Chalas, C. C.; Dunne, M.
Show abstract
In this study we explore the use of protein language models for ranking protein-peptide interaction strength, extending the concept of binary protein interaction classification. We introduce a method that measures and ranks protein binding affinities in an unsupervised manner, eliminating the need for extensive labeled data, structural information, or complex biochemical features. We demonstrate the utility of our approach across five distinct protein-peptide datasets by comparing predicted interaction strength rankings with experimentally derived inhibitory concentration (IC50) values. Furthermore, we discuss limitations encountered during our study and present preliminary findings on extending our approach to more general protein-protein interactions. Finally, we highlight the need for comprehensive datasets specifically designed for ranking protein-protein interactions.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Attention-based approach to predict drug-target interactions across seven target superfamilies 96%
- Patch-DCA: Improved Protein Interface Prediction by utilizing Structural Information and Clustering DCA scores 95%
- Pair-EGRET: enhancing the prediction of protein-proteininteraction sites through graph attention networks and protein language models 95%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.