MechPPI: Binding Mechanism-based Machine-Learning tool for Predicting Protein-Protein Binding Affinity Changes Upon Mutations

Abstract

Protein-protein interactions are essential for various biological processes, including signal transduction, metabolism, vesicle transport, and mitogenic processes. Its crucial to consider them within the context of their interactions with other proteins to understand protein function. Mutations in proteins can affect their binding affinity to partner proteins by introducing various effects, such as changes in hydrophobic regions, electrostatic interactions, or hydrogen bonds. Assessing the impact of mutations on protein interactions can have implications for disease susceptibility and drug efficacy. Understanding the impact of mutations on protein-protein interactions and predicting binding affinity changes computationally can benefit both basic biology and drug development. Different computational methods offer varying levels of accuracy and efficiency, and the choice of method depends on the specific research goals and available resources. We developed MechPPI, a tool that can use potential mechanism features underlying mutation to predict the binding affinity change upon mutation. We showed MechPPI can accurately predict binding affinity change upon a single mutation, and results demonstrate the potential of MechPPI as a powerful and useful computational tool in protein design and engineering.