Advancing network pharmacology with artificial intelligence: the next paradigm in traditional Chinese medicine

Network pharmacology has been widely employed in drug discovery, especially in the research of traditional Chinese medicine (TCM), characterized by the "multi-component, multi-target, and multi-pathway" feature. By incorporating network biology, TCM network pharmacology facilitates the systematic evaluation of therapeutic efficacy and the explicit elucidation of the mechanisms of action, providing a new research paradigm for TCM modernization. The surge of machine learning especially groundbreaking deep learning methods has greatly promoted the development of artificial intelligence (AI) technology, which offers great potential to advance TCM network pharmacology research. Herein, we describe the methodology of TCM network pharmacology, which includes ingredient identification, network construction, network analysis, and experimental validation. We then summarize the key strategies for the construction of various networks and the analysis of constructed networks with AI methods. Finally, we highlight challenges and future directions for cell-cell communication-based network construction, analysis, and validation, offering valuable insights for TCM network pharmacology.