Figure 4. Principal component analysis (PCA) of WT and APP/PS1 mice based on the phenotype of cognitive and pathological character. PCA based on the data of behavioral experiment, LTP recording, immunohistochemical and Nissl staining of WT and APP/PS1 mice (A). Each axis was derived by principal component analysis. Each point represents one of WT and APP/PS1 mice, the number of each point represents month-age of mice. Component 1(variance explained: 90.78%), component 2 (variance explained: 5.97%) considered significant variance with a load below or equal to 0.50 (absolute value). PCA by SAS 9.2 statistics package, the significance level was set at P < 0.05. The mean PCA scores of WT and APP/PS1 mice (B). ***P<0.001, versus the age-matched WT mice, aP<0.05, versus 3-month-old syngeneic mice, bP<0.05, versus 6-month-old syngeneic mice, cP<0.05, versus 9-month-old syngeneic mice by two-way repeated-measures analyses of variance with Tukey multiple comparisons tests. All values are means ± S.D. n=10.