Integrating Large Language Models For Monte Carlo Simulation of Chemical Reaction Networks

TL;DR

This work addresses the tedium of manually defining kinetics for chemical reaction networks by introducing an LLM-driven pipeline that parses natural-language reaction descriptions into kinetic parameters and a stoichiometry matrix suitable for stochastic simulation. By coupling the Gillespie Monte Carlo approach with an evaluator agent and enabling Copasi integration, the framework automates end-to-end simulation workflows from description to results. The approach is validated on multiple published systems, achieving parity with reference results for networks up to 54 reactions and providing Copasi-ready models via Python basico. The findings suggest substantial time savings and broader accessibility for exploring complex CRNs in systems biology, while also highlighting current limitations of open-source LLMs on longer, more detailed inputs.

Abstract

Chemical reaction network is an important method for modeling and exploring complex biological processes, bio-chemical interactions and the behavior of different dynamics in system biology. But, formulating such reaction kinetics takes considerable time. In this paper, we leverage the efficiency of modern large language models to automate the stochastic monte carlo simulation of chemical reaction networks and enable the simulation through the reaction description provided in the form of natural languages. We also integrate this process into widely used simulation tool Copasi to further give the edge and ease to the modelers and researchers. In this work, we show the efficacy and limitations of the modern large language models to parse and create reaction kinetics for modelling complex chemical reaction processes.

Figures (17)

• Figure 1: Example of problem modeled with chemical reaction networks in Amyloid-Beta protein aggregation process (Figure taken from our-1)
• Figure 2: Architecture of Adopted Method
• Figure 3: Results plotted by analyzer step for single species evolution during the chemical reaction process
• Figure 4: Results obtained from the approached methods consisting 54 different reactions, comparable with the original results reported in our-1
• Figure 5: Results obtained from integration of Copasi into the existing workflow, reaction kinetics used are same as in Figure \ref{['fig:evolution-ad-reaction']}