Table 9.
MiR | Parametric p-value | FDR | Permutation p-value | Geom mean of intensities in class 1 | Geom mean of intensities in class 2 | Geom mean of intensities in class 3 | Geom mean of intensities in class 4 |
---|---|---|---|---|---|---|---|
hsa-miR-204-4373094 | 0.00004 | 0.02784 | < 1e-07 | 15.74986 | 11.70222 | 14.82001 | 6.94659 |
hsa-miR-30aSTAR-4373062 | 0.00035 | 0.11658 | 0.00010 | 7.67985 | 6.27768 | 7.25131 | 6.95430 |
The univariate F-test at the nominal significance level of 0.001 was performed among 4 classes: Class 1: Node-negative-Mel 30; Class 2: Node-negative-Mel 60; Class 3: Node-positive-Mel30; Class 4: Node-positive-Mel60. Permutation p-values for significant MiRs were computed based on 10000 random permutations. The Global test: probability of getting at least 2 genes significant by chance (at the 0.001 level) if there are no real differences between the classes was 0.137.