Fig. 6From: Classical and Bayesian random-effects meta-analysis models with sample quality weights in gene expression studiesVenn diagrams present number of differentially expressed (DE) genes in Alzheimer's disease as compared to controls in white matter region using weighted Bayesian random-effects models - Model 1: unweighted BRE with uniform (0,1) model (BRE1), Model 3: unweighted data with Gibb sampling and wP6 weighted between study variance model (BRE3), Model 6: wP6 weighted data with Gibb sampling and wP6 weighted between study variance model (BRE6). This weighted function was applied: \( {w}_{p6}={\left({\sigma}_{ig}^{2\left({w}_{p1}\right)}+{\widehat{\tau}}_g^2\right)}^{-1},{w}_{p1}\in \left\{{2}^{-{S}_{ij}},0.01{\tilde{P}}_{ij}\right\} \), \( {\tilde{P}}_{ij} \) denoted percent of present calls, Sij denoted standardized quality indicators of the jth sample in the ith study. Metadata A: GSE1297, GSE5281, and GSE29378; B: GSE1297, GSE5281, and GSE48350; C: GSE1297, GSE29378, and GSE48350; D: GSE1297, GSE5281, GSE29378, and GSE48350; and E: GSE5281, GSE29378, and GSE48350Back to article page