Deep geometric representations for modeling effects of mutations on protein-protein binding affinity.

Xianggen Liu,Yunan Luo,Sen Song,Pengyong Li,Jian Peng,Roland L Dunbrack

doi:10.1371/journal.pcbi.1009284

Abstract

Modeling the impact of amino acid mutations on protein-protein interaction plays a crucial role in protein engineering and drug design. In this study, we develop GeoPPI, a novel structure-based deep-learning framework to predict the change of binding affinity upon mutations. Based on the three-dimensional structure of a protein, GeoPPI first learns a geometric representation that encodes topology features of the protein structure via a self-supervised learning scheme. These representations are then used as features for training gradient-boosting trees to predict the changes of protein-protein binding affinity upon mutations. We find that GeoPPI is able to learn meaningful features that characterize interactions between atoms in protein structures. In addition, through extensive experiments, we show that GeoPPI achieves new state-of-the-art performance in predicting the binding affinity changes upon both single- and multi-point mutations on six benchmark datasets. Moreover, we show that GeoPPI can accurately estimate the difference of binding affinities between a few recently identified SARS-CoV-2 antibodies and the receptor-binding domain (RBD) of the S protein. These results demonstrate the potential of GeoPPI as a powerful and useful computational tool in protein design and engineering. Our code and datasets are available at: https://github.com/Liuxg16/GeoPPI.

Highlights

Protein-protein interactions (PPIs) play an essential role in many fundamental biological processes
Based on one of these Abs, named C110, GeoPPI locates several residues in its interface where certain mutations can significantly increase the stabilizing effect of the binding with SARS-CoV-2. These results demonstrate that our GeoPPI can serve as a powerful tool for the prediction of binding affinity changes upon mutations and have the potential to be applied in a wide range of tasks, such as designing antibodies with improved binding activity, identifying function-disrupting mutations, and understanding underlying mechanisms of protein biosynthesis
GeoPPI is a deep learning based framework that uses deep geometric representations of protein complexes to model the effects of mutations on the binding affinity

Summary

Introduction

Protein-protein interactions (PPIs) play an essential role in many fundamental biological processes. The antibody (Ab) is a central component of the human immune system that interacts with its target antigen to elicit an immune response. This interaction is performed between the complementary determining regions (CDRs) of the Ab and a specific epitope on the antigen. One of the solutions is to identify affinity-enhancing mutations based on Ab templates [4,5,6]. This strategy faces two-fold challenges when implemented in wet-labs. Fast and inexpensive in silico evaluation of binding affinity changes upon mutations (i.e., ΔΔG) is a promising alternative for screening affinity-enhancing mutations in protein engineering and antibody design

Methods

Results

Discussion

Conclusion