Table 2 Variable selection results from hierarchical spike-and-slab model. The “Component” column gives the module and component number that was selected, the “Cancer” column gives the cancer for which it was selected, “Mean Effect” gives the mean posterior draw for the coefficient of the selected covariate, “Posterior Inclusion Probability” gives the average of inclusion indicators, and “Credible Interval” gives 95% credible interval for coefficient effect
From: A hierarchical spike-and-slab model for pan-cancer survival using pan-omic data
Component | Cancer | Mean effect | Credible interval | Posterior inclusion probability | |
|---|---|---|---|---|---|
1 | 16.1 | UCEC | – 0.50 | (– 0.723, – 0.273) | 1.00 |
2 | 11.1 | KIRC | 0.36 | (– 0.004, 0.678) | 0.92 |
3 | 7.2 | LGG | – 0.47 | (– 0.773, 0.004) | 0.92 |
4 | 14.1 | KIRC | – 0.30 | (– 0.59, 0.005) | 0.88 |
5 | 12.3 | LGG | – 0.29 | (– 0.551, 0.008) | 0.86 |
6 | 5.1 | UCEC | – 0.35 | (– 0.62, 0.006) | 0.86 |
7 | 34.1 | HNSC | – 0.33 | (– 0.599, 0.007) | 0.84 |
8 | 22.1 | KIRP | 0.77 | (– 0.011, 1.435) | 0.82 |
9 | 24.1 | KIRP | – 0.62 | (– 1.218, 0.012) | 0.79 |
10 | 37.1 | ACC | 0.54 | (– 0.013, 1.228) | 0.79 |
11 | 22.1 | KICH | 0.53 | (– 0.012, 1.191) | 0.78 |
12 | 11.1 | KIRP | 0.36 | (– 0.016, 0.964) | 0.77 |
13 | 1.1 | SARC | 0.30 | (– 0.012, 0.662) | 0.74 |
14 | 20.1 | DLBC | 0.54 | (– 0.017, 1.628) | 0.70 |
15 | 37.1 | TGCT | 0.49 | (– 0.017, 1.436) | 0.68 |
16 | 20.1 | BLCA | – 0.21 | (– 0.516, 0.012) | 0.68 |
17 | 11.1 | LIHC | 0.21 | (– 0.079, 0.646) | 0.66 |
18 | 1.1 | LIHC | 0.25 | (– 0.014, 0.659) | 0.63 |
19 | 1.1 | SKCM | 0.18 | (– 0.014, 0.473) | 0.63 |
20 | 20.1 | CORE | – 0.15 | (– 0.434, 0.015) | 0.57 |
21 | 11.1 | CHOL | 0.09 | (– 0.34, 0.673) | 0.54 |
22 | 22.1 | KIRC | 0.12 | (– 0.015, 0.452) | 0.51 |
23 | 39.1 | THYM | – 0.28 | (– 1.052, 0.018) | 0.51 |
24 | 2.1 | PCPG | 0.45 | (– 0.016, 1.642) | 0.50 |