UPLC-QTOFMS based metabolomics followed by stepwise partial least square-discriminant analysis (PLS-DA) explore the possible relation between the variations in secondary metabolites and the phylogenetic divergences of the genus Panax

Phylogenetic and metabolomic approaches have long been employed to study evolutionary relationships among plants. Nonetheless, few studies have examined the difference in metabolites within a clade and between clades of the phylogenetic tree. We attempted to relate phylogenetic studies to metabolomics using stepwise partial least squares-discriminant analysis (PLS-DA) for the genus Panax. Samples were analyzed by ultra-performance liquid chromatography-quadrupole time of flight mass spectrometry (UPLC-QTOFMS) to obtain metabolite profiles. Initially, conventional principal component analysis was subsequently applied to the metabolomic data to show the limitations in relating the expression of metabolites to divisions in the phylogenetic tree. Thereafter, we introduced stepwise PLS-DA with optimized scaling methods, which were properly applied according to the branches of the phylogenetic tree of the four species. Our approach highlighted metabolites of interest by elucidating the directions and degrees of metabolic alterations in each clade of the phylogenetic tree. The results revealed the relationship between metabolic changes in the genus Panax and its species' evolutionary adaptations to different climates. We believe our method will be useful to help understand the metabolite-evolution relationship.