PySAGES, Enhanced Sampling Molecular Dynamics Simulations on GPUs*

When performing molecular dynamics simulations, there exists a large gap between the time scales that can be probed computationally to the ones observed in experiments. To tackle this issue two strategies are commonly used: 1) algorithms that explore the simulated configurational space more efficiently; or 2) hardware accelerators such as GPUs. We combine both in PySAGES (Python Suite for Advanced General Ensemble Simulations), which can be hooked to different molecular dynamics (MD) simulation packages, allowing the user to perform enhanced sampling simulations through a uniform interface without sacrificing the efficiency of the underlying MD implementation (at the time of writing we provide support for HOOMD-blue and OpenMM). The library, is a Python implementation of SSAGES with support for GPUs. Having a Python frontend provides the user flexibility and the ability to easily extend it. We will discuss its features, advantages, technical aspects of the implementation, and present some examples and benchmarks.