Table 1 |
||||
| 30 top-ranked KEGG pathways using a prior probability of 0.05 | ||||
| KEGG pathway | Number of genes | Posterior probability of being included in the model | Odds ratio | Variance per gene (10^-3) |
| ABC transporters | 24 | 1 | Infinity | 4.24 |
| Lysosome | 90 | 1 | Infinity | 3.56 |
| Proteasome | 40 | 1 | Infinity | 3.53 |
| Complement and coagulation cascades | 53 | 1 | Infinity | 3.13 |
| RIG-I-like receptor signaling pathway | 44 | 1 | Infinity | 2.80 |
| ECM-receptor interaction | 63 | 1 | Infinity | 2.59 |
| Cell adhesion molecules (CAMs) | 95 | 1 | Infinity | 2.50 |
| Axon guidance | 79 | 1 | Infinity | 2.46 |
| SNARE interactions in vesicular transport | 31 | 1 | Infinity | 2.39 |
| RNA degradation | 49 | 1 | Infinity | 2.39 |
| Ubiquitin mediated proteolysis | 105 | 1 | Infinity | 2.34 |
| Neuroactive ligand-receptor interaction | 125 | 1 | Infinity | 2.27 |
| PPAR signaling pathway | 58 | 1 | Infinity | 2.16 |
| Ribosome | 76 | 1 | Infinity | 2.15 |
| MAPK signaling pathway | 179 | 1 | Infinity | 2.11 |
| Aminoacyl-tRNA biosynthesis | 34 | 1 | Infinity | 2.08 |
| Endocytosis | 139 | 1 | Infinity | 2.08 |
| Fc gamma R-mediated phagocytosis | 70 | 1 | Infinity | 2.00 |
| Insulin signaling pathway | 97 | 1 | Infinity | 1.88 |
| Cell cycle | 96 | 1 | Infinity | 1.85 |
| Notch signaling pathway | 31 | 1 | Infinity | 1.73 |
| Cytokine-cytokine receptor interaction | 115 | 1 | Infinity | 1.69 |
| Chemokine signaling pathway | 138 | 1 | Infinity | 1.58 |
| Metabolic pathways | 793 | 1 | Infinity | 1.47 |
| Tight junction | 97 | 0.938 | 288 | 1.11 |
| Purine metabolism | 117 | 0.927 | 240 | 1.16 |
| Chronic myeloid leukemia | 58 | 0.919 | 215 | 1.27 |
| Pathways in cancer | 227 | 0.847 | 105 | 0.945 |
| Basal transcription factors | 23 | 0.826 | 90.3 | 1.28 |
| Circadian rhythm - mammal | 5 | 0.590 | 27.4 | 1.77 |
“Odds ratio” indicates the odds ratio between the prior and posterior probability of being included in the model. “Variance per gene” means the estimated variance of the t-statistic per gene. The gene sets are ranked primarily according to the posterior probabilities and secondarily according to the variance per gene.
Skarman et al. BMC Bioinformatics 2012 13:73 doi:10.1186/1471-2105-13-73
