From: Bayesian meta-analysis models for microarray data: a comparative study
Two-Study Simulation Data | True Integration-Driven Discovery Rate, γ= 0.95 | True Integration-Driven Revision Rate, γ= 0.95 | True Genes, pe FDR = 5% | ||||
---|---|---|---|---|---|---|---|
Prior Degrees of Freedom | Â | Â | Â | Â | Â | Â | |
|
| SEI | PI | SEI | PI | SEI | PI |
3 | 3 | 10.8% | 16.2% | 15.9% | 4.5% | 185 | 211 |
6 | 6 | 4.4% | 11.9% | 11.6% | 1.2% | 197 | 216 |
10 | 10 | 3.6% | 10.2% | 7.9% | 0.6% | 202 | 218 |
20 | 20 | 1.7% | 8.1% | 4.4% | 0.5% | 206 | 219 |
40 | 40 | 1.6% | 6.6% | 5.9% | 0.5% | 208 | 221 |
Five-Study Simulation Data | Â | Â | Â | Â | Â | Â | |
3 | 3 | 2.2% | 15.2% | 10.2% | 3.6% | 212 | 241 |
6 | 6 | 0% | 8.8% | 13.0% | 2.8% | 219 | 246 |
10 | 10 | 0% | 6.9% | 13.1% | 2.7% | 225 | 248 |
20 | 20 | 0% | 3.9% | 14.4% | 3.1% | 228 | 250 |
40 | 40 | 0% | 5.1% | 12.8% | 1.8% | 227 | 250 |