AI Agents for Topological Polymer Simulations and Small-Angle Scattering Analysis

We present two innovative AI-agent systems, SasAgent and ToPolyAgent, that leverage large language models (LLMs) and multi-agent frameworks to automate complex workflows in small-angle scattering (SAS) data analysis and coarse-grained molecular dynamics (MD) simulations of topological polymers. SasAgent, built on the SasView software, features a coordinator agent that interprets user text prompts and delegates tasks to specialized agents for scattering length density (SLD) calculation, synthetic data generation, and experimental data fitting. These agents utilize LLM-friendly tools derived from SasView, such as SLD calculators and bump fitting tools, accessible via a user-friendly Gradio interface. SasAgent streamlines SAS research by enabling precise, automated analysis of complex datasets. Complementarily, ToPolyAgent automates coarse-grained MD simulations for topological polymers (linear, ring, brush, star, dendrimer) using a CrewAI-driven pipeline. It processes user prompts to generate polymer configurations, execute LAMMPS simulations with flexible thermostats (Langevin/Nose-Hoover), and compute metrics like radius of gyration, diffusivity, persistence length, end-to-end distance, and polymer form factor. Both systems demonstrate robust error handling, modularity, and extensibility, with ToPolyAgent supporting diverse topologies and solvent conditions. Through integrated examples, we showcase how these LLM-driven agents enhance scientific workflows, offering scalable, prompt-driven solutions for materials characterization and simulation. These advancements highlight the transformative potential of AI in accelerating polymer and soft matter research.