Model ID | Statistical Test Used | Model Assumptions | Outcomes Measured | Notes |
M1 | Limma (Linear Models for Microarray Data) | Normally distributed residuals, linear relationships | Log-fold changes, adjusted p-values | Widely used for small sample sizes; accounts for multiple testing using an empirical Bayes approach. |
M2 | DESeq2 (Differential gene expression analysis based on the negative binomial distribution) | Count data follows a negative binomial distribution | Base mean expression, log2 fold changes, p-values | Suitable for RNA-Seq data; uses shrinkage estimation for dispersions and fold changes. |
M3 | EdgeR (Empirical Analysis of Digital Gene Expression Data in R) | Negative binomially distributed counts, tag wise dispersions | Common dispersion, tagwise dispersion, exact p-values | Optimized for gene expression comparisons with complex experimental designs. |
M4 | t-test (Independent two-sample t-test) | Normally distributed data, equal variances | Mean expression differences, t-statistics, p-values | Simple comparative analysis; less robust to variance in small sample sizes without equal variances. |