NNP/MM: Accelerating Molecular Dynamics Simulations with Machine Learning Potentials and Molecular Mechanics

• DOI: 10.1021/acs.jcim.3c00773
• Publication Date: 2023-09-11T11:26:40Z
• Authors: Raimondas Galvelis; Alejandro Varela-...; Stefan Doerr; Roberto Fino; Peter Eastman; Thomas E. Markland; John D. Chodera; Gianni De Fabritiis
• Keywords: Metadynamics; Molecular mechanics

Machine learning potentials have emerged as a means to enhance the accuracy of biomolecular simulations. However, their application is constrained by significant computational cost arising from vast number parameters compared with traditional molecular mechanics. To tackle this issue, we introduce an optimized implementation hybrid method (NNP/MM), which combines neural network potential (NNP) and mechanics (MM). This approach models portion system, such small molecule, using NNP while employing MM for remaining system boost efficiency. By conducting dynamics (MD) simulations on various protein–ligand complexes metadynamics (MTD) ligand, showcase capabilities our NNP/MM. It has enabled us increase simulation speed ∼5 times achieve combined sampling 1 μs each complex, marking longest ever reported class