A Scalable Molecular Force Field Parameterization Method Based on Density Functional Theory and Quantum-Level Machine Learning

Fast and accurate molecular force field (FF) parameterization is still an unsolved problem. Accurate FFs are not generally available for all molecules, like novel drug-like molecules. While methods based on quantum mechanics (QM) exist to parameterize them with better accuracy, they are computationally expensive and slow, which limits applicability to a small number of molecules. Here, we present Parameterize, an automated FF parameterization method based on neural network potentials, which are trained to predict QM energies. We show, our method produces more accurate parameters than the general AMBER FF (GAFF2), while requiring just a fraction of time compared with an equivalent parameterization using QM calculations. We expect, our method to be of critical importance in computational structure-based drug discovery. Parameterize is available online at PlayMolecule (www.playmolecule.org).