Mid-Level Data Fusion Techniques of LC-MS and HS-GC-MS for Distinguishing Green and Ripe Forsythiae Fructus.

is a crucial plant resource due to its considerable edible and medicinal value. Its fruit, named Forsythiae Fructus (FF), has been widely used in Traditional Chinese Medicine (TCM). According to the fruit maturity stage, FF is categorized into GFF (green Forsythiae Fructus) and RFF (ripe Forsythiae Fructus). In this study, metabolomics based on UPLC-Q/Orbitrap MS and HS-GC-MS, combined with chemometric methods, was employed to differentiate GFF from RFF and identify potential differential metabolites. Additionally, the mid-level data fusion method was employed to integrate data from both techniques, and the performance of the OPLS-DA model (RY = 0.986, Q = 0.974) surpassed that of the single HS-GC-MS technique (RY = 0.968, Q = 0.930). Moreover, using the criteria of VIP > 1 and -value < 0.05, 30 differential compounds were selected via mid-level data fusion, compared to the initial 61 differential compounds identified by single techniques, effectively reducing data noise and eliminating irrelevant variables. This study provides a comprehensive analysis of volatile and non-volatile compounds in FF, offering valuable insights into quality control and clinical differentiation between GFF and RFF. The findings highlight the potential use of multi-technology metabolomics in the quality control of TCM and offer new perspectives for future research on medicinal plants.