DynaMate: An Autonomous Agent for Protein-Ligand Molecular Dynamics Simulations

Force field-based molecular dynamics (MD) simulations are indispensable for probing the structure, dynamics, and functions of biomolecular systems. Despite their broad utility in drug discovery and protein engineering, the technical complexity of a MD setup remains a major barrier for widespread and efficient usage. Agentic LLMs have demonstrated their capacity to autonomously execute multi-step scientific processes. Despite this progress, these systems have not been used to fully automate protein-ligand MD workflows. Here, we present DynaMate, a modular multi-agent framework that autonomously designs and executes complete MD workflows for both protein and protein-ligand systems, and offers free energy binding affinity calculations with the MM/PB(GB)SA method. The framework integrates dynamic tool use, web search, literature query via PaperQA, and a self-correcting behavior. We evaluate its performance first across fifteen benchmark systems of varying complexity, assessing success rate, efficiency, and adaptability. Second, we confirm its robustness by achieving 67% successful executions of MD simulations on a dataset of 827 diverse protein and protein-ligand systems. Overall, DynaMate reduces the typical multi-hour setup time to minutes and autonomously resolves common runtime errors through iterative reasoning. This automated framework paves the way toward standardized and scalable molecular modeling pipelines for biomolecular and drug design applications.