RESUMEN
Three non-targeted methods, i.e. 1H NMR, LC-HRMS, and HS-SPME/MS-eNose, combined with chemometrics, were used to classify two table grape cultivars (Italia and Victoria) based on five quality levels (5, 4, 3, 2, 1). Grapes at marketable quality levels (5, 4, 3) were also discriminated from non-marketable quality levels (2 and 1). PCA-LDA and PLS-DA were applied, and results showed that, the MS-eNose provided the best results. Specifically, with the Italia table grapes, mean prediction abilities ranging from 87% to 88% and from 98% to 99% were obtained for discrimination amongst the five quality levels and of marketability/non-marketability, respectively. For the cultivar Victoria, mean predictive abilities higher than 99% were achieved for both classifications. Good models were also obtained for both cultivars using NMR and HRMS data, but only for classification by marketability. Satisfying models were further validated by MCCV. Finally, the compounds that contributed the most to the discriminations were identified.
Asunto(s)
Análisis de los Alimentos/métodos , Almacenamiento de Alimentos , Espectroscopía de Protones por Resonancia Magnética/métodos , Vitis/química , Nariz Electrónica/estadística & datos numéricos , Análisis de los Alimentos/estadística & datos numéricos , Calidad de los Alimentos , Análisis de los Mínimos Cuadrados , Espectrometría de Masas/métodos , Espectrometría de Masas/estadística & datos numéricos , Análisis Multivariante , Análisis de Componente Principal , Espectroscopía de Protones por Resonancia Magnética/estadística & datos numéricos , Compuestos Orgánicos Volátiles/análisisRESUMEN
Lentil samples coming from two different countries, i.e. Italy and Canada, were analysed using untargeted 1H NMR fingerprinting in combination with chemometrics in order to build models able to classify them according to their geographical origin. For such aim, Soft Independent Modelling of Class Analogy (SIMCA), k-Nearest Neighbor (k-NN), Principal Component Analysis followed by Linear Discriminant Analysis (PCA-LDA) and Partial Least Squares-Discriminant Analysis (PLS-DA) were applied to the NMR data and the results were compared. The best combination of average recognition (100%) and cross-validation prediction abilities (96.7%) was obtained for the PCA-LDA. All the statistical models were validated both by using a test set and by carrying out a Monte Carlo Cross Validation: the obtained performances were found to be satisfying for all the models, with prediction abilities higher than 95% demonstrating the suitability of the developed methods. Finally, the metabolites that mostly contributed to the lentil discrimination were indicated.