New paper: Ranking protein-protein models with large language models and graph

[maykcaldas]

Paper: Ranking protein-protein models with large language models and graph

Authors: Xiaotong Xu and Alexandre M.J. J. Bonvin

Abstract: Protein-protein interactions (PPIs) are associated with various diseases,including cancer, infections, and neurodegenerative disorders. Obtainingthree-dimensional structural information on these PPIs serves as a foundationto interfere with those or to guide drug design. Various strategies can befollowed to model those complexes, all typically resulting in a large number ofmodels. A challenging step in this process is the identification of good models(near-native PPI conformations) from the large pool of generated models. Toaddress this challenge, we previously developed DeepRank-GNN-esm, a graph-baseddeep learning algorithm for ranking modelled PPI structures harnessing thepower of protein language models. Here, we detail the use of our software withexamples. DeepRank-GNN-esm is freely available athttps://github.com/haddocking/DeepRank-GNN-esm

Link: https://arxiv.org/abs/2407.16375

Reasoning: Reasoning: Let's think step by step in order to determine if the paper is about a language model. We start by examining the title, which mentions "large language models" in the context of ranking protein-protein models. Next, we look at the abstract, which describes the use of a graph-based deep learning algorithm that leverages protein language models to rank protein-protein interaction structures. The focus is on applying language models to a specific domain (protein interactions), but it does involve the use of language models.