The integrator loop of the GROMACS molecular dynamics (MD) package has been rewritten to use a modular approach and introduce Monte Carlo (MC) capability. The modularity aims at making the single instructions as independent as possible, having two great advantages. First, it allows easier recomposition, massively simplifying the implementation of new integrator algorithms. Secondly, it allows to explore new parallelisation strategies, as the dependency of tasks is reduced. The addition of MC to this modular integrator allows to investigate a large amount of new algorithms, including hybrid MC/MD sampling schemes and more stable barostats.<br/><br/>The new framework will be released with the next major version of GROMACS, due early 2020. We are demonstrating new functionality aimed at this release, such as new temperature and pressure control algorithms improving simulation stability, multi-time stepping approaches increasing simulation efficiency, and hybrid MC/MD algorithms showing promise to enhance sampling in condensed simulations.

New integrator framework for GROMACS

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.