Leveraging biosynthetic insight and spectrophotometry for comprehensive phenolic profiling in grape skin by HPLC-DAD

While LC-HRMS is the gold standard for identifying phenolic compounds in grapes, integrated spectrophotometric methods offer a simpler, cost-effective alternative for on-line comprehensive analysis. However, the literature lacks a critical assessment of the feasibility and challenges of this approach. Addressing this, in view of the grape biosynthetic background, we demonstrate that HPLC-DAD, when augmented with full-spectrum UV-Vis data, can achieve comparable on-line discrimination. By training algorithms on subclass-specific features (e.g., absorbance ratios, retention times), our method resolves co-eluting compounds with similar spectra. Using AI to digitize known phenolic spectra from literature, we applied this method to a hydroalcoholic extract from Bidaneh Sefid grape peels. The analysis successfully profiled the phenolic content, revealing a dominance of flavonols, particularly rutin, followed by hydroxycinnamates. This validates the approach as a practical and powerful tool for real-time grape extract analysis.