Application of SPORT algorithm on ATR-FTIR data: A rapid and green tool for the characterization and discrimination of three typical Italian Pecorino cheeses

• Published in: Journal of food composition and analysis Vol. 114; p. 104784
• Main Authors: Di Donato, Francesca, Biancolillo, Alessandra, Foschi, Martina, D’Archivio, Angelo Antonio
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.12.2022
• Subjects: algorithms; Attenuated total reflectance-fourier transformed infrared spectroscopy; cheesemaking; Data fusion; Ewes’ milk cheese; Food analysis; Food composition; Italy; manufacturing; Partial least squares-discriminant analysis; prediction; Sequential and orthogonalized partial least squares; Sequential preprocessing through orthogonalization; spectroscopy; sports; Italy; Ewes’ milk cheese; Attenuated total reflectance-fourier transformed infrared spectroscopy; Sequential preprocessing through orthogonalization; Partial least squares-discriminant analysis; Sequential and orthogonalized partial least squares; Data fusion; Food composition; Food analysis
• ISSN: 0889-1575; 1096-0481
• DOI: 10.1016/j.jfca.2022.104784

Three certified Pecorino cheeses, protected designation of origin (PDO) Pecorino Romano, PDO Pecorino Sardo, and Pecorino di Farindola (Slow Food Presidium specialty), produced in Central Italy were analyzed by attenuated total reflectance-fourier transformed infrared (ATR-FTIR) spectroscopy and classified according to their manufacture and/or different cheesemaking process they underwent. The ATR-FTIR spectra were processed by partial least squares-discriminant analysis (PLS-DA) to be used as benchmark method. Moreover, the multi-block strategy sequential preprocessing through orthogonalization (SPORT), based on sequential and orthogonalized partial least squares (SO-PLS), was applied in order to test whether association of spectra preprocessings could enhance the prediction rates. Eventually, the best results were achieved by the multi-block approach, which allows obtaining an accuracy of 98.33% (corresponding to two misclassified samples over 120) in external validation. [Display omitted] •Italian Pecorino cheeses analysed by Attenuated Total Reflectance-FTIR Spectroscopy.•The aim is to discriminate Pecorino cheeses according to three diverse origins.•SPORT was successfully applied to ATR-FTIR spectroscopic data.•SPORT correctly classified 98.3% of the test samples (external validation).