Selected-ion flow-tube mass-spectrometry (SIFT-MS) fingerprinting versus chemical profiling for geographic traceability of Moroccan Argan oils. (15th October 2018)

• Record Type: Journal Article
• Authors: Kharbach, Mourad; Kamal, Rabie; Mansouri, Mohammed Alaoui; Marmouzi, Ilias; Viaene, Johan; Cherrah, Yahia; Alaoui, Katim; Vercammen, Joeri; Bouklouze, Abdelaziz; Vander Heyden, Yvan
• Is Part Of: Food chemistry. Volume 263(2018)
• Page Start: 8
• Page End: 17
• Publication Date: 2018-10-15
• Subjects: γ-Tocopherol (PubChem CID: 92729) -- δ-Tocopherol (PubChem CID: 92094) -- Linoleic acid (PubChem CID: 5280450) -- Oleic acid (PubChem CID: 445639) -- Palmitic acid (PubChem CID: 985) -- Stearic acid (PubChem CID: 5281) -- Schottenol (PubChem CID: 441837) -- Stigma-8–22-dien-3β-ol (PubChem CID: 5280794) -- Spinasterol (PubChem CID: 5281331) -- Δ−7-avenasterol (PubChem CID: 12795736); Argan oil -- Selected-ion flow-tube mass spectrometry -- Geographical origin -- Chemometric class-modeling -- Classification methods -- Fingerprints; Food -- Analysis -- Periodicals; Food -- Composition -- Periodicals; 664
• DOI: 10.1016/j.foodchem.2018.04.059
• Languages: English
• ISSNs: 0308-8146

Abstract

Graphical abstract: Highlights: SIFT-MS fingerprints were recorded for geographical classification of Argan oil. Five classification techniques were evaluated for geographical discrimination. Classical chemical profiling was also applied for geographic Argan-oil discrimination. Variable selection was performed to identify potential biomarkers in SIFT-MS data. Abstract: This study investigated the effectiveness of SIFT-MS versus chemical profiling, both coupled to multivariate data analysis, to classify 95 Extra Virgin Argan Oils (EVAO), originating from five Moroccan Argan forest locations. The full scan option of SIFT-MS, is suitable to indicate the geographic origin of EVAO based on the fingerprints obtained using the three chemical ionization precursors (H3 O +, NO + and O2 + ). The chemical profiling (including acidity, peroxide value, spectrophotometric indices, fatty acids, tocopherols- and sterols composition) was also used for classification. Partial least squares discriminant analysis (PLS-DA), soft independent modeling of class analogy (SIMCA), K-nearest neighbors (KNN), and support vector machines (SVM), were compared. The SIFT-MS data were therefore fed to variable-selection methods to find potential biomarkers for classification. The classification models based either on chemical profiling or SIFT-MS data were able to classify the samples with high accuracy. SIFT-MS was found to be advantageous for rapid geographic classification.