Table 9.
Lymphoma cancer data set
| Ns | ||||||
|---|---|---|---|---|---|---|
| 19 | 24 | 64 | 88 | 96 | ||
| I | ORdensity | 100 | 100 | 100 | 100 | 100 |
| limma | 95.24 | 100 | 100 | 100 | - | |
| SAM | 100 | 100 | 100 | - | - | |
| II | A vs B | 13 | 17 | 14 | 17 | - |
| A vs C | 14 | 16 | 42 | 63 | - | |
| B vs C | 17 | 23 | 58 | - | - | |
| III | ORdensity | 10.2 | 13.2 | 40.4 | 61.4 | 65.3 |
| (1.48) | (2.84) | (2.01) | (2.27) | (3.13) | ||
| limma | 7.9 | 9.7 | 27.4 | 36.7 | - | |
| (0.87) | (1.57) | (2.17) | (1.34) | - | ||
| SAM | 14.2 | 17.3 | 51.4 | - | - | |
| (2.82) | (3.97) | (8.85) | - | - | ||
Results for different number (Ns) of selected genes: 19 with ORdensity relaxed selection; 24 with limma and Bonferroni; 64 with SAM; 88 with limma and BH, and 96 total potential DE genes. Rows I present the leave-one-out cross-validation correct classification rate. In bold, the results for the genes selected under standard criteria for ORdensity, limma and SAM procedures; in rows II, number of common selected genes between the ORdensity, limma and SAM approaches; in rows III, mean and standard deviation (in brackets) of the number of genes that for 10 subsamples were always kept selected