arXiv：cs.AI· Zerui Yang, Yuwei Wan, Siyu Yan, Yudai Matsuda, Tong Xie, Linqi Song·· 7 小时前AI 评分33
DrugMCTS：结合多智能体、RAG 与蒙特卡洛树搜索的药物重定位框架
DrugMCTS: a drug repurposing framework combining multi-agent, RAG and Monte Carlo Tree Search
AI 导读
DrugMCTS 是一个将 RAG、多智能体协作与蒙特卡洛树搜索融合的药物重定位框架，由五个专门智能体负责检索和分析分子与蛋白质信息，实现结构化迭代推理。在 DrugBank 和 KIBA 数据集上，其召回率与鲁棒性显著优于通用 LLM 和深度学习基线。
正文
Abstract:Recent advances in large language models have demonstrated considerable potential in scientific domains such as drug repositioning. However, their effectiveness remains constrained when reasoning extends beyond the knowledge acquired during pretraining. Conventional approaches, such as fine-tuning or retrieval-augmented generation, face limitations in either imposing high computational overhead or failing to fully exploit structured scientific data. To overcome these challenges, we propose DrugMCTS, a novel framework that synergistically integrates RAG, multi-agent collaboration, and Monte Carlo Tree Search for drug repositioning. The framework employs five specialized agents tasked with retrieving and analyzing molecular and protein information, thereby enabling structured and iterative reasoning. Extensive experiments on the DrugBank and KIBA datasets demonstrate that DrugMCTS achieves substantially higher recall and robustness compared to both general-purpose LLMs and deep learning baselines. Our results highlight the importance of structured reasoning, agent-based collaboration, and feedback-driven search mechanisms in advancing LLM applications for drug repositioning.
| Subjects: | Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE) |
| Cite as: | arXiv:2507.07426 [cs.AI] |
| (or arXiv:2507.07426v4 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2507.07426 arXiv-issued DOI via DataCite |
Submission history
From: Zerui Yang [view email]
[v1]
Thu, 10 Jul 2025 04:39:55 UTC (950 KB)
[v2]
Sat, 12 Jul 2025 08:20:44 UTC (950 KB)
[v3]
Thu, 31 Jul 2025 13:57:25 UTC (1,726 KB)
[v4]
Tue, 6 Oct 2026 01:27:31 UTC (1,726 KB)
来源：arXiv：cs.AI · arxiv.org