Exploring UPLC-QTOF-MS-based targeted and untargeted approaches for understanding wine mouthfeel: A sensometabolomic...

This study aimed to establish relationships between wine composition and in-mouth sensory properties using a sensometabolomic approach. Forty-two red wines were sensorially assessed and chemically characterised using UPLC-QTOF-MS for targeted and untargeted analyses. Suitable partial least squares regression models were obtained for "dry", "sour", "oily", "prickly", and "unctuous". "Dry" was positively contributed by flavan-3-ols, anthocyanin derivatives (AntD), valine, gallic acid and its ethyl ester, and peptides, and negatively by sulfonated flavan-3-ols, anthocyanin-ethyl-flavan-3-ols, tartaric acid, flavonols (FOL), hydroxycinnamic acids (HA), protocatechuic ethyl ester, and proline. The "sour" model included molecules involved in "dry" and "bitter", ostensibly as a result of cognitive interactions. Derivatives of FOLs, epicatechin gallate, and N-acetyl-glucosamine phosphate contributed positively to "oily", as did vanillic acid, HAs, pyranoanthocyanins, and malvidin-flavan-3-ol derivatives for "prickly", and sugars, glutathione disulfide, AntD, FOL, and one HA for "unctuous". The presented approach offers an interesting tool for deciphering the sensory-active compounds involved in mouthfeel perception.