Discrimination of apples from the surrounding Bohai Bay and the loess plateau: A combined study of ICP-MS and UPLC-Q-TOF-MS based element and metabolite fingerprints

Jianyi Zhang,Ning Ma,Guofeng Xu,Lixue Kuang,Youming Shen
DOI: https://doi.org/10.1016/j.foodchem.2024.140279
2024-11-30
Abstract:Apples are important fruits in China, and their authentication is beneficial for quality control. However, the differentiation between apples from two primary producing regions, the surrounding Bohai Bay (BHB) and the Loess Plateau (LP), has not been well studied. This study used element and metabolite fingerprints combined with mathematical recognition techniques to discriminate between BHB and LP apples. A total of 235 samples were collected from these regions during 2018-2019. The apple element and metabolite profiles were obtained via instrument analysis. Differential elements and metabolites between BHB and LP apples were identified, and linear and nonlinear discriminant models were constructed. Nonlinear models demonstrated higher accuracy and effectiveness in model optimization. The final random forest (RF) model, constructed with 11 elements and 51 metabolites, achieved a training accuracy of 91.51% and a validation accuracy of 98.57%. This study discriminated between BHB and LP apples, providing a foundation for apple authentication.
What problem does this paper attempt to address?