UPLC-Q-TOF/MS-based Untargeted Metabolomics for Discrimination of Navel Oranges from Different Geographical Origins of China

Yuhao Li,Lu Liang,Chunhui Xu,Tianming Yang,Yuanxing Wang
DOI: https://doi.org/10.1016/j.lwt.2020.110382
IF: 6.056
2021-01-01
LWT
Abstract:Navel oranges (Citrus sinensis. Newhall) are widely planted in several major production areas in southern China, with the Gannan region being the most distinctive. In this study, UPLC-Q-TOF/MS was used to discriminate the geographical origin of navel oranges harvested from distinct production areas in China. Results showed that the unsupervised analysis (PCA) explained 30.8% of the total variance, whereas supervised analysis (OPLS-DA) with cross-validated R2Y and Q2 values of 0.991 and 0.914, respectively, can obtain potential metabolites with better discrimination between Gannan and Non-Gannan regions. Forty-two potential differential metabolites with VIP >1.0 and P < 0.05 (T-test) were selected and identified. Three of these metabolites (narirutin 4′-glucoside, sinapic acid and farnesyl acetone) with AUC > 0.7 and P < 0.05 in logistic-regression analysis were included in the establishment of the receiver operating characteristic combined diagnostic analysis. Further, 16 metabolites were used to generate a suitable classification model for the targeted classification of the samples from different areas. These results provide guidance for geographical discrimination of navel oranges, and its application may be extended to improve planting conditions of navel oranges.
What problem does this paper attempt to address?