Analysis of Variance Heterogeneity in Metabolic Traits Among CSS Mice
Mao Ding
2024-10-24
☆ Objective
This study aims to assess and compare the variance and mean effects of metabolic traits between CSS10 and B6 mice, as well as between CSS17 and B6 mice. The metabolic traits include body weight, body length, glucose levels, BMI measured at different time points, and liver weight along with normalized liver weight at the time of sacrifice. The null hypothesis is neither of the two CSS strains mice (CSS10 & CSS17) will differ significantly from the B6 reference strain for any trait, whether considering central tendency or dispersion.
☆ Density Plot
☆ Violin Plots
☆ Skewness & Kurtosis
The Cullen and Frey graphs depict skewness and kurtosis for all traits in the three mouse strains.
☆ Sample Statistics
Cells highlighted in pink indicate that the Jarque-Bera normality test p-value is greater than 0.05, suggesting a normal distribution. If the cells are empty, this indicates a failure of the bimodality test using GmixR. The 0 for “CSS10_BMI_0_day” results from the Jarque-Bera test method, which shows a p-value of 0 when the actual p-value is less than 2.2e-16.
☆ Basic Comparison
☆ Variance Effect
The green-highlighted cells indicate p-values less than 0.05. For the permutation test (indicated by an asterisk in front of the column), we used 1000 permutations. If you observe a value of 0 in these columns, it indicates that the actual observed difference is greater than all the differences obtained from the permutation tests. Consequently, the p-value is less than 1/number_of permutation. In our case, we used 1000 permutations, so the p-value is less than 1/1000.
☆ Mean Effect
The green-highlighted cells indicate p-values less than 0.05. For the permutation test (indicated by an asterisk in front of the column), we used 1000 permutations. If you observe a value of 0 in these columns, it indicates that the actual observed difference is greater than all the differences obtained from the permutation tests. Consequently, the p-value is less than 1/number_of permutation. In our case, we used 1000 permutations, so the p-value is less than 1/1000.