1H NMR chemometric models for classification of Czech wine type and variety. (1st March 2021)

• Record Type: Journal Article
• Authors: Mascellani, Anna; Hoca, Gokce; Babisz, Marek; Krska, Pavel; Kloucek, Pavel; Havlik, Jaroslav
• Is Part Of: Food chemistry. Volume 339(2021)
• Publication Date: 2021-03-01
• Subjects: 1H NMR -- Chemometrics -- Wine classification -- randomForest -- Wine analysis; Isobutanol (PubChem CID: 6560) -- Isopentanol (Compound CID: 31260) -- Isoleucine (Compound CID: 6306) -- Leucine (Compound CID: 6106) -- 2, 3-Butanediol (Compound CID: 262) -- Ethanol (Compound CID: 702) -- Ethyl Acetate (Compound CID: 8857) -- Lactate (Compound CID: 612) -- Alanine (Compound CID: 5950) -- Proline (Compound CID: 145742) -- Acetate (Compound CID: 175) -- Methionine (Compound CID: 876) -- Acetoin (Compound CID: 179) -- Acetoacetate (Compound CID: 6971017) -- Pyruvate (Compound CID: 107735) -- Glutamate (Compound CID: 33032) -- Succinate (Compound CID: 160419) -- Choline (Compound CID: 305) -- Myo-Inositol (Compound CID: 892) -- Methanol (Compound CID: 887) -- Glycerol (Compound CID: 753) -- Fructose (Compound CID: 2723872) -- Tartrate (Compound CID: 3806114) -- Glucose (Compound CID: 5793) -- Turanose (Compound CID: 5460935) -- Uridine (Compound CID: 6029) -- Catechin (Compound CID: 9064) -- Epicatechin (Compound CID: 72276) -- p-Hydroxyphenylacetic acid (Compound CID: 127) -- Tyrosine (Compound CID: 6057) -- Gallate (Compound CID: 370) -- Phenethyl alcohol (Compound CID: 57361413) -- Phenylalanine (Compound CID: 6140) -- Chlorogenate (Compound CID: 1794427) -- Formate (Compound CID: 283) -- Histidin (Compound CID: 6274) -- Trigonellin (Compound CID: 5570); Food -- Analysis -- Periodicals; Food -- Composition -- Periodicals; 664
• DOI: 10.1016/j.foodchem.2020.127852
• Languages: English
• ISSNs: 0308-8146

Abstract

Highlights: We obtained a large dataset of 917 wines of Czech origin rich in metadata. Machine learning showed good performance in type and variety classification. The algorithm distinguished Pinot blanc and Chardonnay, which are often mistakenly interchanged. Wines from the Znojemská and Mikulovská sub-regions could be differentiated. Abstract: 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 1 H 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.