Table 4.
The predictive ability of four basic methods and ELPGV, and the comparison p-value between ELPGV and others in CD, UC, IBD with IBDGC dataset
| Method | CD | UC | IBD | |||
|---|---|---|---|---|---|---|
| Predictive ability | p-value | Predictive ability | p-value | Predictive ability | p-value | |
| ELPGV | 0.4516 ± 0.0065 | — | 0.7920 ± 0.0045 | — | 0.4253 ± 0.0071 | — |
| BayesA | 0.4359 ± 0.0067 | 1.058E−15 | 0.7817 ± 0.0045 | 7.365E−25 | 0.4128 ± 0.0075 | 2.470 E−13 |
| BayesB | 0.4338 ± 0.0063 | 4.932E−13 | 0.7831 ± 0.0047 | 4.845E−16 | 0.4052 ± 0.0067 | 1.911E−14 |
| BayesCπ | 0.4452 ± 0.0063 | 3.938E−07 | 0.7845 ± 0.0047 | 3.878E−13 | 0.4171 ± 0.0068 | 5.272E−10 |
| GBLUP | 0.3692 ± 0.0063 | 3.659E−34 | 0.6687 ± 0.0052 | 3.314E−56 | 0.3455 ± 0.0079 | 9.965E−35 |
ELPGV is the ensemble learning based on BayesA, BayesB, BayesCπ and GBLUP
— Represents no explicit result was found in this method
Crohn disease (CD) and ulcerative colitis disease (UC)