Principal component analysis (PCA) of 102 peaks that represent different metabolites measured in wild-type (WT) and LT transgenic seeds expressing the T-AtCGS gene. PCA is presented as a combination of the first three dimensions, which together comprise 80.7% of the metabolite variance. Each data point represents an independent sample. The data analysis was performed using TMEV software. Component 1 explained 61.7% of the variance, component 2 12.2%, and component 3 6.8%. The combined percentages of the variance are given in brackets.