A Multi-Agent Framework for Automated Coarse-Grained Molecular Dynamics of Polymers

2026-08-06

AI Summary

This work addresses the inefficiency and limited transferability of conventional bottom-up coarse-grained (CG) molecular dynamics modeling, which typically requires laborious, system-specific derivation of effective potentials for each polymer. The authors propose CGMas, a multi-agent framework that integrates large language models (LLMs) with hierarchical self-correction capabilities into the entire CG pipeline. Starting from natural language descriptions, CGMas autonomously constructs all-atom topologies, equilibrates systems, performs CG mapping, derives interaction potentials via Boltzmann inversion, and validates results—all while enforcing physical constraints through LLM-driven reasoning and multi-agent collaboration. Evaluated on 27 homopolymer and copolymer systems, the method achieves density errors below 5% in 22 cases and reduces simulation time per system from 38–88 minutes to approximately one minute, substantially enhancing both efficiency and generalizability.

Abstract

Coarse-grained (CG) molecular dynamics extends polymer simulation beyond the scales accessible to all-atom (AA) methods, but bottom-up CG modeling is laborious. The CG resolution is a design choice, so a transferable parameter set is generally not available and the potentials are derived anew for each polymer mapping. Here we present CGMas, a multi-agent framework that automates topology construction, equilibration, mapping, potential derivation, and validation from a natural-language specification of the polymer and target resolution. A large-language-model (LLM) reasoning agent infers the AA topology from polymer name, while layered self-correction resolves physical errors common to unsaturated, heteroatom-containing, and polar polymers. Downstream agents equilibrate the system, map it onto CG representation, derive potentials through Boltzmann inversion, and benchmark the model against its atomistic reference. CGMas completed all 27 homopolymer and copolymer tasks, matched the AA density to within 5% in 22, and reduced simulation from 38-88 min to 1 min, establishing agentic LLMs as a route to automated polymer coarse-graining.

Problem

• coarse-grained molecular dynamics
• polymer modeling
• transferable parameters
• potential derivation
• molecular simulation

Innovation

• multi-agent framework
• coarse-grained molecular dynamics
• large language model
• automated polymer modeling
• Boltzmann inversion