Recent advances in interpretable machine learning using structure-based protein representations

TL;DR

This paper reviews recent progress in interpretable machine learning methods for structure-based protein representations, highlighting their applications in predicting structures, functions, and interactions to aid drug discovery and protein engineering.

Contribution

It provides a comprehensive overview of methods for representing protein structures and emphasizes the importance of interpretability in ML models for structural biology.

Findings

Interpretable ML supports protein structure and function prediction.

Structure-based representations enhance understanding of protein interactions.

Interpretable approaches accelerate drug development and protein design.

Abstract

Recent advancements in machine learning (ML) are transforming the field of structural biology. For example, AlphaFold, a groundbreaking neural network for protein structure prediction, has been widely adopted by researchers. The availability of easy-to-use interfaces and interpretable outcomes from the neural network architecture, such as the confidence scores used to color the predicted structures, have made AlphaFold accessible even to non-ML experts. In this paper, we present various methods for representing protein 3D structures from low- to high-resolution, and show how interpretable ML methods can support tasks such as predicting protein structures, protein function, and protein-protein interactions. This survey also emphasizes the significance of interpreting and visualizing ML-based inference for structure-based protein representations that enhance interpretability and knowledge discovery. Developing such interpretable approaches promises to further accelerate fields including drug development and protein design.