mdtools: Tools for Running MD Simulations in OpenMM

This package is intended to provide a useful framework for running MD simulations using OpenMM. The goal of mdtools is to extend OpenMM's funcitonality by facilitating common use cases with sensible default parameter values. With mdtools, you can run an MD simulation of a protein in a waterbox or the unit cell from a crystal structure with just a few lines of Python code.

For now, this repo is very much targeted to my current use cases, which is centered around comparing MD simulations to crystallographic data.

Installation

Pre-requisites to install via conda

The best course of action is to start off by installing openmm, openmmforcefields, and openff-toolkit via conda.

conda install -c omnia -c conda-forge openmm openmmforcefields openff-toolkit

GPU specific installation

If you plan to use openmm/mdtools with GPU acceleration through CUDA (and you probably do!!) you'll need some extra goodies. As long as you run the above conda install command in the same GPU-containing environment that you plan to use for simulations, everything should be installed correctly automagically.

pip-based installation of mdtools and other dependencies

I have not yet made this package available on PyPI. Until then, mdtools can be installed via git+ syntax:

pip install git+https://github.com/Hekstra-Lab/mdtools.git

or, you can clone the repo and install your local copy:

git clone https://github.com/Hekstra-Lab/mdtools.git
cd mdtools
pip install -e .

Note: building the package could be very slow in vanilla Conda due to the many dependencies that need to be resolved. Please consider installing boa and use conda mambabuild for an immense speed-up. Also, use mamba install instead of conda install for additional speed-up.

Examples

Example scripts that make use of the provided libraries will be found in the examples directory (Stay tuned).

Changelog

• Added and reformatted many docstrings for extensive documentation purpose.
• Added many utility functions in mdtools.utils for performing post-simulation analyses on crystal MD systems.
• Support saving velocities and other data through additional reporters in MDSystem.buildSimulation.
• Support reporting force overflows at specific atoms in MDSystem.equilibrate and MDSystem.simulate, which are very helpful for debugging forcefield issues during pilot MD simulation runs.
• Support bulk addition of specified amount of water during squeeze runs and set target volume to reach convergence more rapidly under prior knowledge about system density or volume.
• Several minor QoL changes.