Table 3.
Performance on test set | Performance on validation set | |||||
---|---|---|---|---|---|---|
AUC20 | AUC50 | AUC100 | AUC20 | AUC50 | AUC100 | |
CIPHER_SP | 0.0029* | 0.0046* | 0.0066* | 0 | 0 | 0 |
CIPHER_DN | 0.0015* | 0.0027* | 0.0042* | 0 | 0 | 0 |
RWR | 0.0075* | 0.0178* | 0.0283* | 0.0233 | 0.0358 | 0.0475 |
DK | 0.0192* | 0.0255* | 0.0294* | 0.0211 | 0.0306 | 0.0399 |
RWRH | 0.0916* | 0.1250* | 0.1664* | 0.2009 | 0.2724 | 0.3288 |
MINProp | 0.0771* | 0.1266* | 0.1799* | 0.1963 | 0.2625 | 0.3104 |
BiRW | 0.0421* | 0.0780* | 0.1142* | 0.1544 | 0.2180 | 0.26672 |
PRINCE | 0.1117 | 0.1468 | 0.2088 | 0.1433 | 0.2137 | 0.2715 |
IDLP-G | 0.0040* | 0.0076* | 0.0166* | 0.0189 | 0.0348 | 0.0519 |
IDLP-P | 0.1051* | 0.1457 | 0.1897 | 0.2003 | 0.2592 | 0.3010 |
IDLP | 0.1123 | 0.1492 | 0.1909 | 0.2004 | 0.2572 | 0.2990 |
We compared AUCs when the number of false positive genes are up to 20, 50, 100
*indicates IDLP significantly outperforms the baseline with p<0.05 using Student t-test