Summary

GROMACS Copilot leverages Large Language Models (LLMs) to automate Molecular Dynamics (MD) simulations using GROMACS. It targets researchers and power users, simplifying MD workflows from system setup and execution to analysis by translating natural language prompts into GROMACS commands, thereby saving significant time and manual effort.

How It Works

The agent integrates with LLMs like DeepSeek, OpenAI's GPT-4o, and Gemini. Users input natural language prompts for simulation tasks (e.g., "setup system," "run production MD"). The LLM interprets these, generates GROMACS commands, and executes them within a specified workspace. This abstracts complex command-line operations and parameter tuning, enhancing MD accessibility.

Quick Start & Requirements

• Primary install: pip install git+https://github.com/ChatMol/gromacs_copilot.git
• Prerequisites: conda install -c conda-forge acpype (for protein-ligand complexes), conda install -c conda-forge gmx_mmpbsa (for MM-PBSA/GBSA analysis). Requires API keys for LLM providers (DeepSeek, OpenAI, Gemini) and a working directory with an input PDB file.
• Usage: Invoked via CLI, specifying --workspace, --prompt, --api-key, --model, and --url. An "agent mode" is available for automated long-running trajectories.

Highlighted Details

• Automates end-to-end MD simulation workflows for proteins and protein-ligand complexes.
• Performs automated analysis of RMSD, RMSF, Radius of Gyration (Rg), and Hydrogen Bonds.
• Features an "agent mode" for continuous simulation trajectories and analysis.
• Supports integration with multiple LLM APIs for flexible model selection.

Maintenance & Community

The README provides no details on community channels, notable contributors, sponsorships, or a public roadmap.

Licensing & Compatibility

Dual-licensed: GPLv3 for open-source use, and a separate Commercial License for proprietary applications. GPLv3 imposes copyleft requirements.

Limitations & Caveats

LLMs may select incorrect GROMACS group indices, requiring manual verification. LLM-GROMACS prompt interaction can be imperfect. All results require independent validation; the tool is provided "as is" for research and educational purposes.