Graph Nets for Partial Charge Prediction

2019/09/17 by Wang, Yuanqing, Fass, Josh, Stern, Chaya D. +2
#Chemical Physics (physics.chem-ph) #Computational Physics (physics.comp-ph) #FOS: Computer and information sciences #FOS: Physical sciences #Machine Learning (cs.LG)

paper · doi:10.48550/arxiv.1909.07903

Abstract

Atomic partial charges are crucial parameters for Molecular Dynamics (MD) simulations, molecular mechanics calculations, and virtual screening, as they determine the electrostatic contributions to interaction energies. Current methods for calculating partial charges, however, are either slow and scale poorly with molecular size (quantum chemical methods) or unreliable (empirical methods). Here, we present a new charge derivation method based on Graph Nets---a set of update and aggregate functions that operate on molecular topologies and propagate information thereon---that could approximate charges derived from Density Functional Theory (DFT) calculations with high accuracy and an over 500-fold speed up.

Related

• Chemical Properties from Graph Neural Network-Predicted Electron Densities
• EAC-Net: Predicting real-space charge density via equivariant atomic contributions
• A Recipe for Charge Density Prediction
• A graph neural network charge model targeting accurate electrostatic properties of organic molecules
• DeepDFT: Neural Message Passing Network for Accurate Charge Density Prediction