Assessment of premium organic orange juices authenticity using HPLC-HR-MS and HS-SPME-GC-MS combining data fusion and chemometrics

This study aims to develop a robust chemometric approach to make it possible to authenticate European premium organic orange juices. The metabolomic fingerprinting and the volatile profile of commercial orange juices were analyzed by HPLC-HR-MS and HS-SPME-GC-MS. These data were used for authentication (classification of orange juices) purposes, using principal component analysis, hierarchical cluster analysis and partial least squares discriminant analysis, which provided acceptable results. Some flavonoids, fatty acids, aldehydes and esters were identified as potential markers involved in the differentiation of organic juices. Data fusion strategies were tested and 'mid-level' data fusion achieved an optimal model for classifying organic or conventional orange juices with a sensitivity and specificity of 100%, thus improving the individual models. This approach, combining mass spectrometry techniques, chemometrics and data fusion, likely provides a new framework for the authentication of organic foodstuffs.