Skip to main content
Fig. 6 | BMC Bioinformatics

Fig. 6

From: Identifying novel associations in GWAS by hierarchical Bayesian latent variable detection of differentially misclassified phenotypes

Fig. 6

PheLEx discoveries for bipolar disorder phenotype. a Heatmap showing r2 measure of linkage disequilibrium (LD) computed between the two differentially significant candidate SNPs (rows) identified for corrected bipolar disorder phenotype and training SNPs used as input for PheLEx (columns). Differentially significant SNPs between original and corrected bipolar disorder phenotype with r2 < k, k ~ 1e− 2 (not in LD with training SNPs) are identified as PheLEx discoveries. b Log transformed p-values (y-axis) are reported for a subset of SNPs not in LD with training SNPs plotted in genomic position (x-axis). The small dot for each SNP denotes -log10 (p-value) obtained from association analysis with original bipolar disorder phenotype and large dot denotes -log10 (p-value) obtained from association analysis with the PheLEx corrected bipolar disorder phenotype. Differentially significant SNPs (adjusted p-value < 0.1) in the PheLEx corrected bipolar disorder phenotype compared to original bipolar disorder phenotype are represented in red

Back to article page