Chromatographic Fingerprinting of Chinese Grape Berries

Abstract

Chromatographic fingerprints from three varieties of grapes produced in China (Giant Rose, Red Globe and Summer Black), were obtained by gas chromatography coupled with mass spectrometry. These grapes were subjected to three different production treatments. Two pattern recognition techniques, PCA and PLS-DA were employed to verify the possibility of the creation of a model suitable for the classification of these samples. By means of PCA was verified that the samples could be decomposed according to the grape variety. Also, the varieties of the grapes could be discriminated by the means of PLS-DA (PLS2). Moreover, from PLS-DA (PLS1) models from the Red Globe" variety samples it was verified that it is possible to classify these samples according to the one of the treatments (C), and some trends were observed on the classification of the others (A and B). The other varieties ( Giant Rose" and Summer Black") couldn't be studied as regards the treatments due to the low number of samples.