posted on 2023-05-16, 08:04authored byDessislava Gerginova, Svetlana Simova
Distinguishing the
botanical and geographical origin of wine is
important to prevent wine adulteration and to determine its quality.
The combined use of 1H NMR profiling and chemometrics allows
the quantification of 31 common organic components in the NMR spectra
of 70 wines from different sources. Using the NMR metabolomics approach,
a successful differentiation of wines produced from Bulgarian and
international grape varieties is achieved using linear discriminant
analysis. Wines produced from typical local grape varieties contain
higher average amounts of galacturonic, malic, tartaric, and succinic
acid, alanine, choline, several alcohols, and saccharides arabinose,
galactose, and sucrose than imported wine assortments. A practical
decision tree is proposed for distinguishing 15 different grape varieties
based on the amounts of the common wine components. An example of
distinction of real from diluted wine via creation of a PLS-DA model
is presented. Wines from the two subregions officially recognized
by the EU at the Protected Geographical Indication (PGI) level are
unequivocally recognized.