Exploring a public database to evaluate consumer preference and aroma profile of lager beers by comprehensive two-dimensional gas chromatography and partial least squares regression discriminant analysis

• Published in: Journal of Chromatography A Vol. 1630; p. 461529
• Main Authors: Paiva, Andre Cunha, Hantao, Leandro Wang
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 25-10-2020
• ISSN: 0021-9673
• DOI: 10.1016/j.chroma.2020.461529

•Data dependent analysis of beers using a public database of consumer preference.•U-PLS-DA classification of beer profiles based on their preference rating.•Consumers’ trends seems to be related to a more diversified VOC profile.•Opportunities to combine GC × GC-MS and machine learning in food industry.
In this paper is reported a proof of concept study to evaluate the usage of a public metadata base about beers to guide chemical interpretation of volatile organic compounds (VOC) profiling. 1,569,641 consumers’ evaluations were collected from Untappd® platform and used to define a property of interest according to beer preference. 14 brands of beers from lager family were divided in two groups, first one containing samples with low consumers’ ratings and the second with brands that exhibited high evaluations. VOC profiles were extracted by headspace solid phase microextraction (HS-SPME) and analyzed using comprehensive two-dimensional gas chromatography coupled to mass spectrometry (GC × GC-MS). To correlate the VOC profile and consumers’ preference, unfolded-partial least squares discriminant analysis (U-PLS-DA) with orthogonal signal correction (OSC) were employed. The mathematical model successfully classified all the beer samples. Furthermore, a template match protocol identified 31 compounds related to consumers’ preference. This proof of concept paper revealed the potential usage of public metadata bases for comprehensive chemical interpretation of VOC profiling in foodomics.