PotentialNet for Molecular Property Prediction

TL;DR

PotentialNet employs graph convolutional neural networks to predict molecular properties relevant to drug discovery, achieving state-of-the-art results and introducing new evaluation metrics and validation strategies for chemical data.

Contribution

The paper introduces PotentialNet, a novel graph convolutional neural network architecture tailored for molecular property prediction, with improved performance and new validation methods.

Findings

State-of-the-art performance in protein-ligand binding affinity prediction

New metric for early enrichment in chemical data

Validation strategy based on structural homology clustering

Abstract

The arc of drug discovery entails a multiparameter optimization problem spanning vast length scales. They key parameters range from solubility (angstroms) to protein-ligand binding (nanometers) to in vivo toxicity (meters). Through feature learning---instead of feature engineering---deep neural networks promise to outperform both traditional physics-based and knowledge-based machine learning models for predicting molecular properties pertinent to drug discovery. To this end, we present the PotentialNet family of graph convolutions. These models are specifically designed for and achieve state-of-the-art performance for protein-ligand binding affinity. We further validate these deep neural networks by setting new standards of performance in several ligand-based tasks. In parallel, we introduce a new metric, the Regression Enrichment Factor $EF_\chi^{(R)}$, to measure the early enrichment of computational models for chemical data. Finally, we introduce a cross-validation strategy based on structural homology clustering that can more accurately measure model generalizability, which crucially distinguishes the aims of machine learning for drug discovery from standard machine learning tasks.