Biomarkers-based classification between green teas and decaffeinated green teas using gas chromatography mass spectrometer coupled with in-tube extraction (ITEX). (15th January 2019)

• Record Type: Journal Article
• Authors: Zhang, Lihua; Ku, Kang-Mo
• Is Part Of: Food chemistry. Volume 271(2019)
• Page Start: 450
• Page End: 456
• Publication Date: 2019-01-15
• Subjects: Green tea and decaffeinated green tea -- Volatile organic compounds -- Biomarker -- Random forest; Food -- Analysis -- Periodicals; Food -- Composition -- Periodicals; 664
• DOI: 10.1016/j.foodchem.2018.07.137
• Languages: English
• ISSNs: 0308-8146

Abstract

Highlights: Green tea flavors were analyzed by GC-MS coupled with in-tube extraction (ITEX). ITEX parameters were optimized by central composite design. Random forest classification was used to select the biomarkers. 4 discriminant biomarkers were identified between green tea and decaffeinated tea. Abstract: For identifying discriminatory biomarkers between green tea (GT) and decaffeinated green tea (dGT), in-tube extraction (ITEX)-gas chromatograph-mass spectrometer (GC-MS) was optimized to determine volatile organic compounds (VOCs) from tea products. Biomarker selection between GT and dGT was then conducted by random forest (RF). Optimized ITEX parameters by central composite design were an incubation temperature of 92 °C, incubation time of 12 mins, and 36 for syringe strokes. A training group of 24 samples and testing group of 21 samples were used to RF classification of biomarkers identification. Results revealed that 2-hexenal, 2-ethyl furan, indole, and β-ocimene were selected as discriminatory biomarkers between GT and dGT in the training group. Using these biomarkers with RF classification algorithms, prediction accuracy for dGT and GT were 88.9% and 100%, respectively, which was higher than for other classification algorithms. This implies that ITEX-GC-MS can be a promising tool for quality control of commercial tea products.