Skip to main content
Figure 5 | BMC Bioinformatics

Figure 5

From: Genome-wide estimation of gender differences in the gene expression of human livers: Statistical design and analysis

Figure 5

Expected classification errors for genes that are selected based upon their p-value from the t-test for a gender difference. When the genes are ordered by their p-values from the t-test of a gender difference, the horizontal axis indicates the set of genes selected as having a gender difference. The vertical axis plots associated estimates of the false discovery rate (solid blue line), false non-discovery rate (dash-dotted green line), and proportion of affected genes not selected (dashed red line). These estimates assume that 8% of the 2800 expressed genes, 224 genes, have a gender difference where the 8% was estimated by fitting a mixture of distributions model to the observed p-values.

Back to article page