Back

MPBind: Multitask Protein Binding Site Prediction by Protein Language Models and Equivariant Graph Neural Networks

wang, y.; Boadu, F.; Cheng, J.

2025-04-18 bioinformatics
10.1101/2025.04.12.648527 bioRxiv
Show abstract

Proteins interact with a variety of molecules, including other proteins, DNAs, RNAs, ligands, ions, and lipids. These interactions play a crucial role in cellular communication, metabolic regulation, immune response, and structural integrity, making proteins fundamental to nearly all biological functions. Accurately predicting protein binding (interaction) sites is essential for understanding protein interactions and functions. Here, we introduce MPBind, a multitask protein binding site prediction method, which integrates protein language models (PLMs) that can extract structural and functional information from sequences and equivariant graph neural networks (EGNNs) that can effectively capture geometric features of 3D protein structures. Through multitask learning, it can predict binding sites on proteins that interact with five key categories of binding partners: proteins, DNA/RNA, ligands, lipids, and ions, which is more comprehensive than most existing task-specific methods that can predict only one or a few kinds of binding sites. Moreover, MPBind outperforms both general bind site prediction methods and task-specific binding site prediction methods.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"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.