Skip to main content
Fig. 1 | BMC Biology

Fig. 1

From: RNA-Seq in 296 phased trios provides a high-resolution map of genomic imprinting

Fig. 1

Miami and quantile-quantile plots of genome-wide results for parentally biased gene expression in 165 lymphoblastoid cell lines (LCL) and 131 whole blood (WB) samples. All data shown are based on bidirectional RNA-Seq data. In both a LCLs and b whole blood, two statistical tests for parental bias were used: the upper panel in each cohort shows the results from the paired Wilcoxon signed-rank test, and the lower panel shows the results from the ShrinkBayes test. −log10 transformed adjusted p values are shown on the y-axis and chromosome and position on the x-axis: the dotted green lines indicate a statistical threshold of 10% FDR, with all genes exceeding this highlighted and labeled according to their paternal expression ratio and number of informative samples (see legend). These plots show the results of the analysis based on known transcript annotations, and thus do not include the unannotated transcript at 13q21.2 identified by sliding window analysis. c, e QQ plots for the paired Wilcoxon signed-rank test in LCLs and whole blood. d, f QQ plots for ShrinkBayes in LCLs and whole blood. Note for ShrinkBayes, some of the observed –log10 p values are infinite, indicated by an asterisk on the y-axis. In each plot, the top 30 genes are highlighted and colored according to their paternal ratio. For both cell cohorts and statistical tests, the genomic inflation factor is approximately equal to 1. For genes with multiple UGFs (Additional file 3), we only plot data for the UGF with the most significant p value

Back to article page