Wine flavor and quality are determined by the assessment of multiple sensory stimuli, including aroma, taste, and mouthfeel. It is therefore important to consider the contribution of as many metabolites as possible when attempting to relate wine composition to quality. In this study, partial least squares regression of the volatile (untargeted headspace solid-phase microextraction coupled with gas chromatography time-of-flight mass spectrometry), non-volatile (untargeted reverse-phase ultra-high-performance liquid chromatography mass spectrometry), and combined metabolite profiles were used to predict Pinot Noir wine quality ratings as assessed by experts. Non-volatile metabolite profiles predicted wine quality ratings better than volatile metabolite profiles, suggesting that the non-volatile composition of Pinot Noir wines contributes to quality perception to a greater extent than the volatile composition. This was underscored by descriptive sensory analysis, which found that taste and mouthfeel attributes were better correlated with wine quality ratings than aroma attributes. Important predictors of Pinot Noir wine quality were also characterized. Some new relationships between wine metabolites and quality ratings were found: dipeptides and unsaturated fatty acids were positively related to Pinot Noir wine quality, while N-(3-methylbutyl)acetamide and xanthine were negatively associated.
Keywords: Pinot Noir; metabolomics; non-volatiles; partial least squares regression; volatiles; wine quality.