Key points are not available for this paper at this time.
BACKGROUND: Traditional Chinese Medicine (TCM) presents a unique therapeutic paradigm characterized by multi-compound, multi-target interventions, yet this complexity impedes mechanistic understanding and standardization. While artificial intelligence (AI) has been applied to isolated aspects of TCM research, a critical gap remains in integrating these applications across the inherent hierarchical structure of TCM-from the pharmacological effects of single compounds (SC) to the synergistic mechanisms of complex Chinese medicinal materials (CMM) and Chinese medicine formulae (CMF). AIM OF REVIEW: This review aims to introduce a novel, AI-driven framework that unifies the multi-scale target analysis continuum of TCM through a systematic, cross-scale data flow, positioning AI as the central catalyst for a holistic understanding across SC, CMM, and CMF levels. KEY SCIENTIFIC CONCEPTS OF REVIEW: The proposed framework demonstrates how molecular targets predicted at the SC level serve as foundational inputs to decipher multi-SC synergistic networks within CMM. These modular networks are subsequently integrated to unravel the complex multi-target synergistic mechanisms of CMF, thereby paving the way for intelligent CMF recommendation (CMFR) and precise quantitative dosage prediction. Furthermore, the review critically addresses fundamental challenges such as the "semantic gap" between abstract TCM theories and molecular data, strongly advocating for a "computation-experiment" closed loop to validate in silico predictions. Finally, we propose transformative future directions, including the development of TCM-specific large language models (LLMs), to decode TCM's pharmacological logic and chart a definitive path towards its scientific validation and global integration.
Hong et al. (Sun,) studied this question.