Simple identification of discriminative markers for four Citrus species using a combination of molecular networking and multivariate analysis

Ion-fragmentation information from high-resolution mass spectrometry (HRMS) has gained prominence as chemometric data for analyzing the molecular distribution and changes in food-ingredient metabolites. However, efficient analytical approaches are required to analyze the numerous HRMS signals. Molecular networking (MN) methods have been developed to generate networks between metabolites, based on pattern similarities in their mass spectra. Here, MN was combined with in silico annotation methods and multivariate analysis to identify six discriminative markers for four Citrus samples (C. unshiu mature peel, C. unshiu immature peel, C. aurantium immature fruits, and Poncirus trifoliata immature fruits), namely, naringin, hesperidin, neohesperidin, poncirin, nobiletin, and tangeretin. The proposed markers were experimentally verified and quantitatively monitored based on the species and maturity by HPLC-UV analysis. Neohesperidin and poncirin were identified as discriminative markers for the immature fruits of C. aurantium and Poncirus trifoliata, whereas hesperidin and nobiletin could distinguish the mature peel of C. unshiu from the immature peel.

Automated summary

Ion-fragmentation information from HRMS is used as chemometric data for analyzing molecular distribution and changes in food-ingredient metabolites. In this study, molecular networking (MN) methods combined with in silico annotation methods and multivariate analysis were used to identify six discriminative markers for four Citrus samples. These markers were experimentally verified and quantitatively monitored based on their species and maturity.