SchNet - a deep learning architecture for molecules and materials

TL;DR

SchNet is a deep learning architecture designed for modeling atomistic systems, accurately predicting molecular and material properties, and enabling efficient molecular dynamics simulations and quantum-mechanical studies.

Contribution

The paper introduces SchNet, a novel deep learning model utilizing continuous-filter convolutional layers for atomistic systems, advancing property prediction and molecular dynamics simulations.

Findings

Accurately predicts molecular and material properties across chemical space.

Learns chemically plausible embeddings of atom types.

Enables quantum-mechanical property predictions for complex molecules.

Abstract

Deep learning has led to a paradigm shift in artificial intelligence, including web, text and image search, speech recognition, as well as bioinformatics, with growing impact in chemical physics. Machine learning in general and deep learning in particular is ideally suited for representing quantum-mechanical interactions, enabling to model nonlinear potential-energy surfaces or enhancing the exploration of chemical compound space. Here we present the deep learning architecture SchNet that is specifically designed to model atomistic systems by making use of continuous-filter convolutional layers. We demonstrate the capabilities of SchNet by accurately predicting a range of properties across chemical space for \emph{molecules and materials} where our model learns chemically plausible embeddings of atom types across the periodic table. Finally, we employ SchNet to predict potential-energy surfaces and energy-conserving force fields for molecular dynamics simulations of small molecules and perform an exemplary study of the quantum-mechanical properties of C$_{20}$-fullerene that would have been infeasible with regular ab initio molecular dynamics.