Systematic Characterization of Tea Varieties and Manufacturing Processes through Chemical Profiling and Multivariate Statistical Analysis

The extensive diversity of tea, resulting from varietal traits and manufacturing processes, presents challenges for classification. This study systematically investigated the effects of tea varieties and processing methods on phytochemical profiles. Results revealed substantial variability in caffeoylquinic acid (CQA) profiles within raw and bitter raw Pu'er teas, both of which were characterized by distinct enrichment of 4-caffeoylquinic acid (4-CQA) or 3-O-caffeoylquinic acid (3-CQA), reflecting their diverse chemical signatures. Notably, the differential stability of teaghrelins (TGs) and strictinin (ST) during processing stages underscores their potential as analytical markers and highlights the metabolic divergence driven by fermentation processes. Furthermore, the model robustness and discriminative performance were validated through external prediction, demonstrating the reliable classification of tea extracts. This study provides a comprehensive framework that integrates metabolite profiling with multivariate classification modeling to assess chemical diversity and offer insights applicable to the tea industry.