Boosting LLM's Molecular Structure Elucidation with Knowledge Enhanced Tree Search Reasoning

2025-06-28

AI Summary

Large language models (LLMs) suffer from insufficient domain-specific chemical knowledge, leading to poor accuracy in spectral-to-structure mapping for molecular structure elucidation. To address this, we propose K-MSE, a knowledge-augmented reasoning framework comprising three core components: (1) a curated molecular substructure knowledge base enabling external knowledge retrieval; (2) a differentiable molecular–spectrum matching scoring model that serves as a reward signal for structural hypothesis evaluation; and (3) Monte Carlo Tree Search (MCTS) for test-time dynamic knowledge expansion and efficient exploration of the structural hypothesis space. Evaluated on GPT-4o-mini and GPT-4o, K-MSE achieves over 20% absolute improvement in elucidation accuracy compared to baseline LLM-based approaches, demonstrating substantial gains in robustness and generalization. The framework significantly outperforms prior methods without requiring model retraining or fine-tuning. Our implementation is publicly available.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Knowledge Representation and Reasoning: Knowledge Acquisition
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information

Abstract

Molecular structure elucidation involves deducing a molecule's structure from various types of spectral data, which is crucial in chemical experimental analysis. While large language models (LLMs) have shown remarkable proficiency in analyzing and reasoning through complex tasks, they still encounter substantial challenges in molecular structure elucidation. We identify that these challenges largely stem from LLMs' limited grasp of specialized chemical knowledge. In this work, we introduce a Knowledge-enhanced reasoning framework for Molecular Structure Elucidation (K-MSE), leveraging Monte Carlo Tree Search for test-time scaling as a plugin. Specifically, we construct an external molecular substructure knowledge base to extend the LLMs' coverage of the chemical structure space. Furthermore, we design a specialized molecule-spectrum scorer to act as a reward model for the reasoning process, addressing the issue of inaccurate solution evaluation in LLMs. Experimental results show that our approach significantly boosts performance, particularly gaining more than 20% improvement on both GPT-4o-mini and GPT-4o. Our code is available at https://github.com/HICAI-ZJU/K-MSE.

Problem

• Enhancing LLMs' molecular structure deduction from spectral data
• Addressing LLMs' limited chemical knowledge in structure elucidation
• Improving accuracy in molecular solution evaluation via knowledge integration

Innovation

• Knowledge-enhanced reasoning framework for molecular elucidation
• Monte Carlo Tree Search for test-time scaling
• External molecular substructure knowledge base extension