Can wine composition predict quality? A metabolomics approach to assessing Pinot noir wine quality as rated by experts

Abstract

The perception of wine quality is determined by the assessment of multiple sensory stimuli, including aroma, taste, mouthfeel and visual aspects. With so many different parameters contributing to the overall perception of wine quality, it is important to consider the contribution of all metabolites in a wine when attempting to relate composition to quality. Presently, links between wine composition and quality are largely anecdotal, with winemakers relying on their experience, refined palates, and well established measures of wine quality such as alcohol content, phenolic composition and the absence of major faults to produce high quality wines. 

In this study, we assessed relationships between wine composition and quality ratings determined by wine experts. Forty-eight Pinot noir wines from two vintages and several geographic regions around the world were subjected to sensory and chemical analysis. A panel of experts made up of wine industry professionals (n = 24) assessed the quality of the wines, as well as a number of other sensory attributes. The wines were analysed by untargeted reverse phase UHPLC-MS, and untargeted HS-SPME-GC-TOF-MS to obtain the non-volatile and volatile profiles of each wine respectively. Partial least squares regression of the non-volatile, volatile and combined chemical profiles, together with ratings of wine quality by experts, showed that the non-volatile profiles were more strongly correlated with perceived wine quality than the volatile profiles. Some new correlations between wine metabolites and quality ratings were found: several dipeptides and unsaturated fatty acids were positively associated with wine quality, and a volatile acetamide was strongly negatively correlated. Both the non-volatile wine matrix and the volatile profile of a wine should be considered in the relationship between Pinot noir wine composition and quality.