MT-Mol:Multi Agent System with Tool-based Reasoning for Molecular Optimization

TL;DR

MT-Mol is a multi-agent system that uses tool-guided reasoning with specialized language model agents to improve molecular optimization, achieving state-of-the-art results on multiple benchmark tasks.

Contribution

The paper introduces MT-Mol, a novel multi-agent framework that integrates comprehensive chemical tools with role-specific LLM agents for interpretable and effective molecular optimization.

Findings

Achieved state-of-the-art performance on 17 out of 23 PMO-1K benchmark tasks.

Demonstrated effective collaboration among specialized agents for chemically grounded reasoning.

Enhanced interpretability through stepwise, tool-aligned molecular optimization process.

Abstract

Large language models (LLMs) have large potential for molecular optimization, as they can gather external chemistry tools and enable collaborative interactions to iteratively refine molecular candidates. However, this potential remains underexplored, particularly in the context of structured reasoning, interpretability, and comprehensive tool-grounded molecular optimization. To address this gap, we introduce MT-Mol, a multi-agent framework for molecular optimization that leverages tool-guided reasoning and role-specialized LLM agents. Our system incorporates comprehensive RDKit tools, categorized into five distinct domains: structural descriptors, electronic and topological features, fragment-based functional groups, molecular representations, and miscellaneous chemical properties. Each category is managed by an expert analyst agent, responsible for extracting task-relevant tools and enabling interpretable, chemically grounded feedback. MT-Mol produces molecules with tool-aligned and stepwise reasoning through the interaction between the analyst agents, a molecule-generating scientist, a reasoning-output verifier, and a reviewer agent. As a result, we show that our framework shows the state-of-the-art performance of the PMO-1K benchmark on 17 out of 23 tasks.