TCMDATA: an integrated R package for target-oriented network pharmacology analysis and AI-assisted interpretation in traditional Chinese medicine

Background Existing network pharmacology workflows for traditional Chinese medicine (TCM) remain fragmented across databases and software tools, with limited support for systematic target prioritization. This study aimed to develop TCMDATA , which is an integrated R package for reproducible, target-oriented TCM network pharmacology analysis. Methods TCMDATA is an integrated R package providing a scriptable, target-oriented workflow for TCM network pharmacology. It integrates herb-compound-target associations from multiple public resources, comprising 1,808 herbs, 2,068 target-linked compounds, and 15,889 human protein targets after data cleaning and gene-symbol standardization. For target prioritization, TCMDATA natively implements 19 PPI centrality metrics, four community detection algorithms, and a configurable network perturbation framework for assessing target importance. A multi-algorithm machine learning module supports consensus candidate selection, while an AI agent module enables task routing and customizable skills for automated analysis and interpretation. Results In a case study of Ganoderma lucidum in sepsis, TCMDATA prioritized four candidate targets, MAPK14 , CASP1 , ARG1 , and PPARG . External disease databases, literature evidence, and molecular docking further supported the relevance of these targets to sepsis, illustrating the utility of TCMDATA as a reproducible analytical resource for TCM network pharmacology. Conclusion TCMDATA provides an integrated and reproducible R framework for target-oriented TCM network pharmacology, supporting transparent data integration, network analysis, and machine learning-based candidate prioritization. Interpretation of its outputs should consider the coverage of available public databases, and further experimental validation will help establish the mechanistic relevance of prioritized targets. TCMDATA is freely available at https://github.com/YuLab-SMU/TCMDATA .