Learning the shape of protein micro-environments with a holographic convolutional neural network

TL;DR

This paper introduces Holographic Convolutional Neural Network (H-CNN), a physically motivated machine learning model that predicts protein function from structure, accurately modeling amino acid preferences and mutation impacts to aid protein design.

Contribution

The paper presents a novel H-CNN model that incorporates physical interactions in proteins, improving prediction of mutation effects and functional mapping from structural data.

Findings

H-CNN accurately predicts mutation impacts on protein stability and binding.

H-CNN reflects physical interactions and evolutionary data in protein modeling.

The model can guide the design of proteins with desired functions.

Abstract

Proteins play a central role in biology from immune recognition to brain activity. While major advances in machine learning have improved our ability to predict protein structure from sequence, determining protein function from structure remains a major challenge. Here, we introduce Holographic Convolutional Neural Network (H-CNN) for proteins, which is a physically motivated machine learning approach to model amino acid preferences in protein structures. H-CNN reflects physical interactions in a protein structure and recapitulates the functional information stored in evolutionary data. H-CNN accurately predicts the impact of mutations on protein function, including stability and binding of protein complexes. Our interpretable computational model for protein structure-function maps could guide design of novel proteins with desired function.