The isoflavone diversity (44 varieties) of the soybean, Glycine max (L.) Merrill, from China, Japan, and Korea was examined by high-performance liquid chromatography. The profiles of 12 isoflavones identified from the grains were subjected to data-mining processes, including partial least-squares discriminant analysis (PLS-DA), Pearson's correlation analysis, and hierarchical clustering analysis (HCA). Although PLS-DA did not reveal significant differences among extracts of soybean from 3 countries, the results clearly show that the variation between varieties was low. The CS02554 variety was separate from the others in the first 2 principal components of PLS-DA. HCA of these phytochemicals resulted in clusters derived from closely related biochemical pathways. Daidzin, genistin, and glycitin contents were significantly correlated with their respective malonyl glycoside contents. Daidzein content correlated positively with genistein content (r=0.8189, P<0.0001). The CS02554 variety appears to be a good candidate for future breeding programs, as it contains high levels of isoflavone compounds. These results demonstrate the use of metabolite profiling combined with chemometrics as a tool for assessing the quality of food and identifying metabolic links in biological systems.
Keywords: Chemometrics; Correlation; Glycine max; Isoflavone; Partial least squares discriminant analysis; Soybean.
Copyright © 2013 Elsevier Ltd. All rights reserved.