Skip to main content
Fig. 3 | BMC Bioinformatics

Fig. 3

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

Fig. 3

Improvement in GWAS performance via PheLEx in simulations. Results are shown for simulated true phenotype (green; no misclassification), simulated misclassified phenotype (blue), and PheLEx corrected phenotype (red). a Receiver Operating Characteristic (ROC) curves are shown with mean Sensitivity (y-axis) and mean 1 - Specificity (x-axis) in identifying disease-associated SNPs using p-values obtained from association analyses. b Box plots of area under ROC curve (AUC ROC) values (y-axis) are shown across increasing misclassification rates (x-axis). c Mean precision (y-axis) over recall (x-axis) curves are shown for identifying disease-associated SNPs using p-values obtained from association analyses. d Box plots of area under Precision-Recall curve (AUC PR) values (y-axis) are shown across increasing misclassification rates (x-axis)

Back to article page