Table 1.
Median AUROC and AUPR for all networks in R6 and R8, as well as the default networks of each, respectively (bold indicates higher number per comparison).
| R6 | R8 | R6 (default) | R8 (default) | |
|---|---|---|---|---|
| GO TERM SIZE = 3–10 | ||||
| Median AUROC | 0.650 | 0.694 | 0.627 | 0.684 |
| 95% CI | ±0.316 | ±0.311 | ±0.334 | ±0.329 |
| versus R6 (p-value) | *4.74 × 10−82 | *1.41 × 10−27 | ||
| versus R8 (p-value) | *8.11 × 10−10 | |||
| GO TERM SIZE = 11–300 | ||||
| Median AUROC | 0.871 | 0.890 | 0.857 | 0.882 |
| 95% CI | ±0.217 | ±0.195 | ±0.246 | ±0.212 |
| versus R6 (p-value) | *9.96 × 10−258 | *5.68 × 10−129 | ||
| versus R8 (p-value) | *4.53 × 10−35 | |||
| GO TERM SIZE = 3–10 | ||||
| Median AUPR | 0.012 | 0.019 | 0.019 | 0.026 |
| 95% CI | ±0.349 | ±0.409 | ±0.343 | ±0.408 |
| versus R6 (p-value) | *1.35 × 10−28 | *1.34 × 10−18 | ||
| versus R8 (p-value) | *5.19 × 10−19 | |||
| GO TERM SIZE = 11–300 | ||||
| Median AUPR | 0.185 | 0.220 | 0.181 | 0.215 |
| 95% CI | ±0.412 | ±0.528 | ±0.415 | ±0.529 |
| versus R6 (p-value) | *8.45 × 10−256 | 4.62 × 10−1 | ||
| versus R8 (p-value) | 3.56 × 10−2 | |||
The Wilcoxon rank sum test was performed on the following pairs conditions: R6 versus R8, R6 versus R6 (default), and R8 versus R8 (default). The p-values for these tests are listed with statistically significant values (p < 0.01) marked with an asterisk.