gmxapi: a high-level interface for advanced control and extension of molecular dynamics simulations

Area of Science:

• Computational biology
• Biophysics
• Molecular modeling

Background:

• Molecular dynamics (MD) simulations are crucial for biomolecular applications.
• Extending MD software with custom interactions and simulation ensembles is often necessary.
• Existing GROMACS capabilities require modifications for advanced use cases.

Purpose of the Study:

• To introduce a Python high-level interface for the GROMACS simulation package.
• To enable custom potential functions and ensemble simulations without altering GROMACS code.
• To facilitate experimentally-driven refinement of protein conformational ensembles.

Main Methods:

• Developed a Python API for GROMACS, maintaining its performance.
• Implemented an abstract interface for building and executing computational graphs.
• Ensured minimal dependencies for portability and maintainability.

Main Results:

• The API allows custom potential functions, integrating experimental data.
• It supports running ensembles of simulations efficiently.
• Demonstrated application in experimentally-driven protein conformational ensemble refinement.

Conclusions:

• The new Python interface simplifies extending GROMACS for advanced biomolecular simulations.
• It offers a portable and maintainable solution for custom interactions and ensemble studies.
• This tool aids in the refinement of protein structures using experimental data.