Geographic Variation and Quality Consistency of Toddaliae Asiaticae Radix: A Hybrid Framework Integrating...

Background: Toddaliae Asiaticae Radix (TA) boasts a long history of medicinal application. However, origin traceability and quality assessment of the widely distributed original plant Toddalia asiatica have not been fully elucidated. Methods: A hybrid framework integrating targeted metabolomics, network pharmacology (NP), and machine learning (ML) was established. By optimizing key parameters, a high-coverage and rapid method for multiple categories compounds was developed using ultra-high performance liquid chromatography-multiple reaction monitoring tandem mass spectrometry (UPLC-MRM MS/MS). Using samples collected across 16 geographical regions, redundancy analysis (RDA) and pattern recognition techniques were applied to explore environment-sensitive metabolites. Taking into account five types of diseases, NP analysis was employed to obtain the bioactive components and their contribution weight in disease treatment. Subsequently, core Quality Markers (Q-Markers) with dual functions of responsive to geographic variations and biologically relevant to therapeutic efficacy were figured out, and were used to establish origin scoring model and discrimination model. Results: The geographical metabolic characteristics of the TA from broad regions in China were thoroughly analyzed, and 60 geographically sensitive compounds were identified. Through NP analysis, 27 core Q-Markers were locked. The bioactivity-weighted scoring model based on Q-Markers revealed the consistency of regional rankings as well as minor fluctuations across five diseases. ML demonstrated that the Q-Markers preserved regional discrimination performance, and the introducing of practical-oriented weights enhanced overall discriminative confidence. Conclusions: This research decodes the Geographical metabolic characteristics of TA, and highlights the necessity of function-oriented prioritization of drug resources.