Table 1 Table comparing the type I error of the joint score test statistic, U ψ with tissue-by-tissue (TBT) analysis, MetaTissue (MT) model (FE = Fixed Effects model; RE = Random Effects model) and multivariate Bayesian Model Averaging. Note that all the results are based on 5,000 simulations on 100 observations at a nominal level of α=0.05
From: Exploiting expression patterns across multiple tissues to map expression quantitative trait loci
| Â | Number of tissues = 5 | Number of tissues = 10 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
MAF | TBT | M T(F E) | MT(RE) | BMA | U ψ | TBT | M T(F E) | MT(RE) | BMA | U ψ |
0.05 | 0.0422 | 0.0410 | 0.0488 | 0.0488 | 0.0456 | 0.0434 | 0.0362 | 0.0404 | 0.0392 | 0.0416 |
0.10 | 0.0476 | 0.0442 | 0.0510 | 0.0488 | 0.0494 | 0.0512 | 0.0368 | 0.0472 | 0.0448 | 0.0480 |