Gen.parRep: A first implementation of the Generalized Parallel Replica dynamics for the long time simulation of metastable biochemical systems

Abstract

Metastability is one of the major encountered obstacles when performing long molecular dynamics simulations, and many methods were developed to address this challenge. The “Parallel Replica”(ParRep) dynamics is known for allowing to simulate very long trajectories of metastable Langevin dynamics in the materials science community, but it relies on assumptions that can hardly be transposed to the world of biochemical simulations. The later developed “Generalized ParRep” variant solves those issues, but it was not applied to significant systems of interest so far.In this article, we present the program gen.parRep, the first publicly available implementation of the Generalized Parallel Replica method (BSD 3-Clause license), targeting frequently encountered metastable biochemical systems, such as conformational equilibria or dissociation of protein–ligand complexes. It will be shown that the resulting C++ implementation exhibits a strong linear scalability, providing up to 70% of the maximum possible speedup on several hundreds of CPUs. Program summaryProgram Title: gen.parRepProgram Files doi:http://dx.doi.org/10.17632/dgp8wdgtgt.1Licensing provisions: BSD 3-clauseProgramming language: C++ (mostly), C and LuaNature of problem: Molecular dynamics simulations of chemical and biological systems usually encounter the problem of metastability, because of the timescale separation between the time discretization step used for dynamics and the usual mean time between conformational changes. The use of Accelerated dynamics [1] methods is usually necessary in order to address this challenge.Solution method: The Generalized Parallel Replica method [2] accelerates the exit from metastable states, providing a linear speedup of N, N being the number of replicas of the system running in parallel. This C++ implementation, the first available so far, exhibits a strong linear scaling on hundreds of CPUs, therefore ready for production studies on High Performance Computing (HPC) machines.Additional comments: Git repository: https://gitlab.inria.fr/parallel-replica/gen.parRep