Geographical verification of Pleuropterus multiflorus thunb. by functional compounds, stable isotope ratios, and multielement combined with machine learning methods

The dry roots of Pleuropterus multiflorus Thunb. (PM) have been traditionally utilized as functional foods and medicines China and various Asian countries. They are extensively cultivated in multiple provinces in China, with variations in prices and qualities. This study aims to investigate the regional characteristics of PM by 4 stable isotope ratios, 40 multielement and 16 functional compounds contents, using a total of 357 samples from 8 different geographical origins. Machine learning methods were developed to authenticating the geographical origins of PM, yielding the accuracy range from 94.44 % to 100 % in the test set. Notably, the protected designation of origin, Deqing PM, exhibited a high accuracy of 100 % in most models, A total of 30 significant prediction variables, encompassing 16 functional compounds, δ2H, 12 rare earth elements, and Cu, were identified. Additionally, the study identified altitude, high temperature, and dry or moisture index as the primary influenced environmental factors.