Rapid evaluation of antioxidant activity of Rheum tanguticum: A synergistic strategy of near-infrared spectroscopy, chromatographic effects, and machine learning

This study integrates HPLC fingerprinting and NIR spectroscopy to evaluate the antioxidant capacity of Rheum tanguticum. HPLC fingerprinting comprehensively characterized the complex chemical composition, identifying 13 compounds. Meanwhile, NIR spectroscopy provided a rapid, non-destructive approach to predict antioxidant capacity. The combination of these techniques offers a holistic and efficient strategy for evaluating antioxidant activity. The spectrum-effect relationships were analyzed by correlating the results of three antioxidant assays (ABTS, DPPH, and FRAP) with the peak areas of the common peaks identified by HPLC. Glycosylated anthraquinones, particularly conjugated forms like rhein-8-O-glucoside, were found to synergistically enhance antioxidant activity with free aglycones. NIR models optimized using the PLS method with Python-based algorithms demonstrated strong predictive abilities, achieving RPD values of 2.43 for ABTS, 2.63 for DPPH, and 2.43 for FRAP. This research provides a rapid method to evaluate the antioxidant capacity of Rh. tanguticum, offering a reference for future applications in other fields.