Construction of a GNPS-based activity molecular network coupled with 1H NMR HetCA: A strategy for rapid antioxidant identification from Aganosma cymosa

The rapid and targeted identification of bioactive constituents from complex natural product matrices remains a significant challenge in natural product research. This study employs an integrated biochemometrics workflow combining 1 H Nuclear Magnetic Resonance ( 1 H NMR) and mass spectrometry (MS) to efficiently pinpoint antioxidant components from Aganosma cymosa . Systematic separation of the crude extract yielded 37 microfractions (MFs), which were subjected to both antioxidant activity evaluation and UHPLC-MS/MS profiling. An active molecular network was constructed by calculating Spearman rank correlation coefficients between MS feature abundances and the corresponding antioxidant activity data, enabling visualization of activity-correlated features within the GNPS platform. To address the limitation of MS-based annotation in distinguishing structural isomers, feature-based GNPS analysis was integrated with 1 H NMR heterospectral covariance analysis (HetCA). The HetCA pseudo-spectra provided orthogonal validation and precise spectroscopic targets for subsequent isolation. Through this combined strategy, the main antioxidant constituents were successfully predicted, isolated, and structurally verified by NMR data. In conclusion, this study demonstrates that the synergistic integration of Spearman rank correlation-based active molecular networking, a multi-level prioritization strategy, and 1 H NMR HetCA orthogonal validation provides a powerful paradigm for the precise identification of bioactive components in complex plant matrices.