arXiv 2606.013942026-05-31cs.CL

UniD$^3$：一种用于药物-疾病发现与推理的知识图谱增强RAG框架

UniD$^3$: A Knowledge Graph-Enhanced RAG Framework for Drug-Disease Discovery and Reasoning

Department of Health Outcomes and Biomedical Informatics, University of Florida（佛罗里达大学健康成果与生物医学信息学系）；Department of Hematology, H. Lee Moffitt Cancer Center and Research Institute（血液科，H. Lee Moffitt癌症中心与研究院）；Center for Bioinformatics and Computational Biology, Houston Methodist Research Institute（生物信息学与计算生物学中心，休斯顿方法主义研究学院）；Department of Cardiothoracic Surgery, Weill Cornell Medicine, Cornell University（心胸外科，Weill Cornell医学，康奈尔大学）；Department of Biostatistics and Health Data Science, Indiana University School of Medicine（生物统计学与健康数据科学系，印第安纳大学医学院）

Qing Wang, Tianshi Liu, Minghao Zhou, Jialu Liang, Sen Guo, Guangyu Wang, Jing Su, Qianqian Song

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出UniD$^3$框架，结合大语言模型与知识图谱增强检索生成（KG-RAG），从生物医学文献中提取、组织和验证药物-疾病知识，生成结构化数据集并提升推理可靠性。

摘要：

药物-疾病关系的系统表征对于药物发现和重定位至关重要，但受限于生物医学文献的异质性和快速增长。现有数据集依赖劳动密集型整理且常不完整，而仅使用大语言模型的方法存在幻觉和证据基础薄弱的问题。我们提出UniD$^3$，一个统一框架，将大语言模型与知识图谱增强检索生成（KG-RAG）相结合，在药物-疾病匹配（DDM）、药物有效性评估（DEA）和药物-靶点分析（DTA）中提取、组织和验证药物-疾病知识。UniD$^3$使用Llama 3.3-70B处理157,849篇PubMed文章，并通过双阶段策略构建知识图谱，该策略结合了论文级提取和以药物和疾病实体为中心的KG级整合。这些图谱支持基于KG-RAG的结构化数据集生成，并通过外部基准测试、与精选资源的模糊匹配以及临床医生评审进行评估。UniD$^3$生成了六个知识图谱和大规模数据集，包括28,915个DDM、15,042个DEA和超过4,000个DTA问答对。外部验证显示性能强劲（DDM/DEA的F1为0.85-0.87；DTA为0.82），临床医生评审确认高可靠性（AUROC = 0.90）。KG-RAG增强模型优于独立大语言模型，UniD$^3$聊天机器人支持可解释、有引用的药物-疾病关系探索。UniD$^3$提供了一个可扩展、可扩展的框架，用于将非结构化生物医学文献转化为高质量、结构化的药物-疾病知识，支持AI驱动的发现、重定位和精准医学。

英文摘要：

Systematic characterization of drug-disease relationships is essential for drug discovery and repurposing, yet is hindered by the heterogeneity and rapid growth of biomedical literature. Existing datasets rely on labor-intensive curation and are often incomplete, while LLM-only approaches suffer from hallucination and weak evidence grounding. We introduce UniD$^3$, a unified framework that integrates Large Language Models with Knowledge Graph-enhanced Retrieval-Augmented Generation (KG-RAG) to extract, organize, and validate drug-disease knowledge across Drug-Disease Matching (DDM), Drug Effectiveness Assessment (DEA), and Drug-Target Analysis (DTA). UniD$^3$ processes 157,849 PubMed articles with Llama 3.3-70B and constructs knowledge graphs via a dual-stage strategy combining paper-level extraction with KG-level consolidation centered on drug and disease entities. These graphs support KG-RAG-based generation of structured datasets, evaluated through external benchmarks, fuzzy matching with curated resources, and clinician review. UniD$^3$ produces six knowledge graphs and large-scale datasets, including 28,915 DDM, 15,042 DEA, and over 4,000 DTA QA pairs. External validation shows strong performance (F1: 0.85-0.87 for DDM/DEA; 0.82 for DTA), with clinician review confirming high reliability (AUROC = 0.90). KG-RAG-augmented models outperform standalone LLMs, and the UniD$^3$ chatbot enables interpretable, citation-supported exploration of drug-disease relationships. UniD$^3$ provides a scalable, extensible framework for transforming unstructured biomedical literature into high-quality, structured drug-disease knowledge, supporting AI-driven discovery, repurposing, and precision medicine.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。