Table 2. Number of differentially expressed genes by different statistical tests for Dataset 1, where #G up, #G dw denote up and down-regulated genes, respectively. Here, Pearson’s correlation test can not be used as the number of experimental samples is not equal to the number of control samples.
| parametric tests at 0.0001 p-value cutoff | non-parametric tests at 0.0001 p-value cutoff | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| t-test | Welch’s t-test | Bayes’ t-test | common | Limma | SAM | Wilcoxon | permute | common | |
| #G up | 616 | 586 | 376 | 344 | 115 | 136 | 188 | 176 | 93 |
| #G dw | 619 | 642 | 481 | 403 | 325 | 387 | 615 | 582 | 320 |