AI驱动下中药化学成分的智能分析与化学空间拓展OA

The compositional analysis of complex material systems in traditional Chinese medicine(TCM)—including small molecules,proteins,polysaccharides,and higher-order self-assembled structures—has long been constrained by insufficient spectral library coverage and the bottleneck of manual interpretation,thereby limiting the standardization and depth of research into their material basis.Leveraging its strong capability for multi-source data integration and deep representation learn-ing,artificial intelligence(AI)is driving a paradigm shift in TCM constituent analysis from experi-ence-driven approaches to data-driven methodologies.This review systematically surveyed learning paradigms,core architectures,and representative tasks of AI-based TCM constituent analysis,with a particular focus on molecular representation learning methods,spectroscopic data processing strate-gies,and recent advances in complex multidimensional component identification and chemical space expansion.Finally,key challenges and future directions were discussed,including data standard-ization,multimodal integration,and model interpretability.From an interdisciplinary perspective,this review aimed to provide methodological support for the modernization and high-quality develop-ment of traditional Chinese medicine.