This presentation from September 2021 provides a comprehensive introduction to GROMACS. By solving Newton’s equations of motion, GROMACS enables the sampling of a molecular system’s phase space to extract accurate physical and thermodynamic properties. The presentation details the essential simulation workflow—ranging from initial system preparation (structure, topology, and parameter files) to energy minimization, equilibration, and production runs—while highlighting GROMACS’ compatibility with major force fields and its optimization for CPU and GPU architectures. The session explores specialized applications such as density-guided simulations for cryo-EM data and the use of restraints for NMR-based structure determination.

This tutorial provides step-by-step instructions on how to run a simple molecular dynamics simulation of a small protein in water using GROMACS. The tutorial is provided as an interactive Jupyter notebook which you can run in your browser or it can be downloaded to run locally on your machine.

Access the complete GROMACS documentation including links to version downloads, release notes, user guide and more!

The ask.bioexcel GROMACS category is meant for all GROMACS-related topics. From software installation to usage, development, and more!

Access a number of GROMACS tutorials ranging from basic to advanced applications and including guidance on how to navigate the tutorial suite. All tutorials are provided via interactive Jupyter notebooks. The page includes links to other GROMACS training resources.

This presentation provides a strategic overview of how to achieve optimal performance in GROMACS by effectively mapping simulation tasks to modern hardware. While traditional manual parameter tweaking has been largely superseded by automated features like PME tuning, significant performance gains now rely on understanding the interplay between CPU threads and GPU acceleration. The talk details parallelization layers including OpenMP threads, MPI message passing, and 3D domain decomposition, emphasizing that “more hardware” does not always equate to better performance due to communication overheads. For many systems, maximizing total scientific throughput—often by running multiple independent simulations per GPU using the -multi or -gpu_id options—is more efficient than attempting to accelerate a single run across all available resources.

This presentation provides a technical overview of optimizing molecular dynamics simulations through the precise configuration of .mdp parameters. The session emphasizes the critical relationship between simulation parameters and force field selection, detailing best practices for integrators, neighbor searching via the Verlet scheme, and the implementation of PME electrostatics and Van der Waals cutoffs. Key recommendations include utilizing the velocity-rescale thermostat for robust temperature coupling and maintaining force field consistency to ensure physical accuracy. By highlighting how to balance computational performance with data output management, the presentation serves as a practical guide for researchers to achieve reliable and reproducible biomolecular simulations using GROMACS.

This presentation introduces the Accelerated Weight Histogram (AWH) method as a powerful tool for overcoming sampling barriers in molecular dynamics simulations. AWH functions as an iterative solver that adaptively develops a bias potential to flatten free-energy landscapes along defined reaction coordinates, enabling the observation of rare events—such as DNA base pairing or molecular permeation—that occur on millisecond timescales. Key advantages of the method include its robust, automated convergence properties, full integration within GROMACS for high performance, and support for “multiple walkers” to achieve super-linear scaling. While the method efficiently handles the sampling problem and calculates accurate free-energy profiles, the presentation emphasizes that the primary challenge for researchers remains the appropriate selection of reaction coordinates for their specific biomolecular systems.

This presentation discusses the application of the Accelerated Weight Histogram (AWH) method to alchemical free energy calculations within GROMACS. By treating the coupling parameter lambda as a dynamic variable and employing an adaptive biasing potential, AWH ensures uniform sampling across the alchemical pathway and eliminates the need for numerous independent, fixed-lambda simulation windows. The AWH method excels in its efficiency, particularly through the use of multiple parallel walkers to accelerate convergence, and its robustness in handling complex reaction coordinates. Demonstrated through practical examples—ranging from simple solvation free energies to intricate lipid barrier permeability studies—AWH is presented as a user-friendly and highly efficient alternative to traditional methods like the Bennett Acceptance Ratio (BAR) for calculating precise free energy differences in molecular systems.