Evaluation of Soybean Varieties for Soy Milk Based on Factor and Cluster Analysis

Abstract

Soy milk is a widely popular soybean product with many advantages. As its raw material, how the quality indexes of soybean affects the quality of soy milk worth to be investigated. In this experiment, 70 representative soybean varieties were selected, and the physicochemical indexes were determined. After the soy milk was prepared from each variety, the indicators were measured, with sensory evaluation assisted. In the results, for the soybean varieties, the average 100-grain weight, water absorption rate, protein content was 23.10 g, 132.90% and 39.21%, respectively. Moreover, the soy milk indexes including the average yield of soybean milk (1.78%), protein content (78.80%), protein transfer rate (26.24%), solid transfer rate (59.33%), and Stickiness (24.07 g) were determined. It was found by correlation analysis that, many of these indexes were correlated. The Principal component analysis (PCA), cluster analysis and factor analysis showed that, the comprehensive scores of the varieties ‘Xinxuan 88’ and ‘Shenxingdou’ were above 85 points. The quality of soybean varieties is related to the density, yield, sugar content and viscosity of soybean milk. To sum up, this study provides a reference for soybean germplasm evaluation and the development of soybean products represented by soy milk through the analysis of soybean indexes and the production indexes.