| Autor: |
Galvelis R; Acellera Labs, C/Doctor Trueta 183, Barcelona 08005, Spain.; Computational Science Laboratory, Universitat Pompeu Fabra, PRBB, C/Doctor Aiguader 88, Barcelona 08003, Spain., Varela-Rial A; Acellera Ltd, Devonshire House 582 Honeypot Lane, Stanmore Middlesex, HA7 1JS, United Kingdom., Doerr S; Acellera Ltd, Devonshire House 582 Honeypot Lane, Stanmore Middlesex, HA7 1JS, United Kingdom., Fino R; Acellera Labs, C/Doctor Trueta 183, Barcelona 08005, Spain., Eastman P; Department of Chemistry, Stanford University, 337 Campus Drive, Stanford, California 94305, United States., Markland TE; Department of Chemistry, Stanford University, 337 Campus Drive, Stanford, California 94305, United States., Chodera JD; Computational and Systems Biology Program, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, New York, New York 10065, United States., De Fabritiis G; Computational Science Laboratory, Universitat Pompeu Fabra, PRBB, C/Doctor Aiguader 88, Barcelona 08003, Spain.; Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig Lluis Companys 23, Barcelona 08010, Spain.; Acellera Ltd, Devonshire House 582 Honeypot Lane, Stanmore Middlesex, HA7 1JS, United Kingdom. |
| Abstrakt: |
Machine learning potentials have emerged as a means to enhance the accuracy of biomolecular simulations. However, their application is constrained by the significant computational cost arising from the vast number of parameters compared with traditional molecular mechanics. To tackle this issue, we introduce an optimized implementation of the hybrid method (NNP/MM), which combines a neural network potential (NNP) and molecular mechanics (MM). This approach models a portion of the system, such as a small molecule, using NNP while employing MM for the remaining system to boost efficiency. By conducting molecular dynamics (MD) simulations on various protein-ligand complexes and metadynamics (MTD) simulations on a ligand, we showcase the capabilities of our implementation of NNP/MM. It has enabled us to increase the simulation speed by ∼5 times and achieve a combined sampling of 1 μs for each complex, marking the longest simulations ever reported for this class of simulations. |
