Table 2.
Wilcoxon rank test for all pairs of training conditions. Each entry in the table is the P-value for testing the hypothesis that the ranks resulting from a pair of experiments come from the same distribution. These are performed on the RNA decoy set. We provide the test results for the DecoyFinder decoy set in Supplementary Table S3 material and show consistent results.
| Experiment 2 | aBPN | Secondary | Primary | No-label | Majority | Swap | Random |
|---|---|---|---|---|---|---|---|
| Experiment 1 | |||||||
| ABPN + unsup. | 2.9e–06 | 5.1e–26 | 1.4e–22 | 2.1e–21 | 9.3e–25 | 7.2e–26 | 2.3e–18 |
| ABPN | – | 1.7e–11 | 5.6e–11 | 1.5e–08 | 4.3e–10 | 6.4e–12 | 2.0e–08 |
| Secondary | – | 3.2e–01 | 7.7e–01 | 1.3e–01 | 2.8e–02 | 1.7e–01 | |
| Primary | – | 4.3e–01 | 2.7e–01 | 2.4e–02 | 3.2e–01 | ||
| No-label | – | 5.5e–01 | 1.5e–02 | 1.8e–01 | |||
| Majority | – | 3.7e–01 | 3.3e–01 | ||||
| Swap | – | 5.5e–01 |