Untargeted lipidomic approach in studying pinot noir wine lipids and predicting wine origin

An untargeted lipidomic profiling approach based on ultra - performance liquid chromatography - time-of-flight tandem mass spectrometry (UPLC-TOF-MS/MS) was successfully used to study the origin of commercial Pinot noir wines. The total wine lipids were extracted using a modified Bligh-Dyer method. In all wine samples, the total lipids were less than 0.1% (w/w) of wine. The wines analyzed consisted of 222 lipids from 11 different classes. 48 commercial Pinot noir wine samples were collected from producers in Burgundy, California, Oregon, and New Zealand. Lipidomic data was studied using advanced multivariate analysis methods, random forest, k-nearest neighbor (k-NN), and linear discriminant analysis. The overall classification accuracy was 97.5% for random forest and 90% for k-NN. Wine lipids showed a strong potential for classifying wines by origin, with the top 58 lipids contributing to the discrimination. This information could potentially be used for further study of the impacts of lipids on wine characteristics and authenticity.

1H NMR chemometric models for classification of Czech wine type and variety

A set of 917 wines of Czech origin were analysed using nuclear magnetic resonance spectroscopy (NMR) with the aim of building and evaluating multivariate statistical models and machine learning methods for the classification of 6 types based on colour and residual sugar content, 13 wine grape varieties and 4 locations based on ¹H NMR spectra. The predictive models afforded greater than 93% correctness for classifying dry and medium dry, medium, and sweet white wines and dry red wines. The trained Random Forest (RF) model classified Pinot noir with 96% correctness, Blaufränkisch 96%, Riesling 92%, Cabernet Sauvignon 77%, Chardonnay 76%, Gewürtztraminer 60%, Hibernal 60%, Grüner Veltliner 52%, Pinot gris 48%, Sauvignon Blanc 45%, and Pálava 40%. Pinot blanc and Chardonnay, varieties that are often mistakenly interchanged, were discriminated with 71% correctness. The findings support chemometrics as a tool for predicting important features in wine, particularly for quality assessment and fraud detection.

1H NMR metabolomics applied to Bordeaux red wines

The q-NMR metabolomics has already demonstrated its potential for classifying wines of different geographical origins, grape varieties, or vintages. This study focuses on the characterisation of Bordeaux red wines, seeking to discriminate them from others produced in the major French wine regions. A sampling of 224 commercial French wines was analysed by ¹H NMR and forty compounds were quantified. Non-supervised and supervised statistical analyses revealed a singular imprint of Bordeaux wines in comparison with other French wines, with classification rates ranging from 71% to 100%. Within the Bordeaux vineyards, red wines from the different Bordeaux subdivisions were analysed from different vintages. Our results indicate that q-NMR metabolomics enables the differentiation of Médoc and Libournais vineyard highlighting the most discriminant constituents. In addition, the effects of wine evolution during bottle aging and vintage on Bordeaux red wines were pointed out and discussed.

Characterization of white wines from north-eastern Italy with surface-enhanced Raman spectroscopy

In the present study, label-free SERS spectroscopy is applied as a useful analytical technique for white wine characterization. 180 samples of three white wines varieties from northeastern Italy, Sauvignon Blanc, Ribolla Gialla and Friulano, collected from three different Italian producers from 2016 vintage, have been analyzed using Ag citrate-reduced colloids and a portable Raman instrument with a 785 nm laser. A PCA of SERS spectra showed that discrimination between wines and wineries is possible. Main spectral differences are due to adenine, carboxylic acids and glutathione, with their ratio changing among different wine types and producers. A robust version of the Soft Independent Modelling of Class Analogy (SIMCA) method was used to model the class space of each wine and to perform the classification among the different categories, yielding overall efficiencies between 87 and 93%. These results are extremely encouraging and open the way to the application of this SERS protocol as a wine identification assay.

Classification of red wines using suitable markers coupled with multivariate statistic analysis

Methodologies for chemometric classification of five authentic red wine varieties from Murfatlar wine center, Romania, young and aged are reported. The discriminant analysis based on several anthocyanins, organic acids, (13)C/(12)C, (18)O/(16)O and D/H isotopic ratios, (1)H and (13)C NMR fingerprints revealed a very satisfactory categorization of the wines, both in terms of variety and vintage, thus illustrating the validity of selected variables for wine authentication purposes. LDA applied to the combined data shows 85.7% classification of wines according to grape variety and 71.1% classification of wines according to vintage year, including a control wine set for each categorization, thus allowing an accurate interpretation of the data. Thereby, anthocyanins, certain anthocyanin ratios, oxalic, shikimic, lactic, citric and succinic acids, sugars like glucose, amino acids like histidine, leucine, isoleucine and alanine, and also 2,3-butanediol, methanol, glycerol and isotopic variables were significant for classification of wines.