Figure 5.

Metabolic pattern recognition and data analysis. ((a) and (b)) PCA score plot 3D and 2D of three groups and QC samples. PC1, first principle component score; PC2, second principle component score; PC3, third principle component score. The more aggregated the QC samples were, the higher the quality of experimental data was. Ellipses represented 95% confidence intervals. (c) PCA score plot 3D (group HD vs M). ((d) and (e)) PLS-DA score plot 2D (group HD vs M; group W vs M). R2Y represented the interpretation rate of the second principal component of the model and Q2Y represented the prediction rate of the model.