Using biomolecular simulations at an atomistic level has become an invaluable tool to complement as well as guide experimental strategies for research in academia and industry. Simulations are used to, for example, shed light on protein interactions and kinetics and can range from estimating free energies of binding of small molecules to a protein, to consolidating observed dynamics in NMR experiments to protein motion found in a simulated ensemble. While there is a wide array of software available to, for example, estimate what the binding free energy of a ligand to a protein is, these often require expert knowledge to use them or come with an expensive license fee. BioSimSpace was designed to make molecular simulations more accessible to non-simulation experts, allowing for the user to focus on a specific scientific question. BioSimSpace wraps around common molecular simulation tools both for creating the simulation trajectory data as well as their analysis. Simulation packages such as Gromacs and AMBER as well as SOMD are supported and trajectories are conveniently returned in either MDtraj or MDAnalysis format. The user can draw from a library of existing — so-called — nodes each written to perform a particular simulation task, e.g. solvating a protein and equilibrating the protein water box. These nodesare written in Python and can either be executed in Python, as a command line script or run as part of a Knime workflow. With the availability of a high-level Python API, users can also customize or write their own nodes from scratch. In this way, using different nodes, a whole complex simulation and analysis workflow can be built up to address questions such as what are the free energies of binding of a given ligand set to a protein. Advantages in using BioSimSpace aside from the ease of use are the automatic creation of easily sharable and reproducible simulation workflows extraneous to the underlying simulation software. This will be demonstrated by showing results from a complex alchemical free energy workflow, taken from the recent D3R grand challenge 4, run with two different simulation engines (SOMD and Gromacs), but executed in an identical fashion in BioSimSpace.

BioSimSpace: Interoperable molecular software framework

2019 Fall National Meeting

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