Fig. 7

Computational discovery and experimental validation of miRNA-mediated tissue-specific edQTL:eQTL colocalization events. a Flowchart of computational analysis. Each edQTL site was required to have at least one tissue with colocalizing edQTL and eQTL signals (PP4 > 0.75) and at least one tissue with non-colocalizing edQTL and eQTL signals (PP1 > 0.75), be in the 3′-UTR, and have an editing-specific miRNA which is differentially expressed (fold change > 2) between the colocalizing and non-colocalizing tissues. b Example of an edQTL event in RPL13 with colocalizing edQTL and eQTL signals in the skin (not sun exposed) (left column) and non-colocalizing edQTL and eQTL signals in the cerebellum (right column). Manhattan plots for edQTL (top row), eQTL (middle row), and scatter plots of −log10(p value) from edQTL and eQTL signals (bottom row) show the presence (left column) or absence (right column) of edQTL:eQTL colocalization. Colocalization posterior probabilities are shown in parentheses (bottom row). c Differential expression of an editing-specific miRNA (miR-26b-5p) targeting the edited version of RPL13. Tissues expressing high levels of the miRNA have colocalizing edQTL and eQTL signals. Tissues expressing low levels of the miRNA do not have colocalizing edQTL and eQTL signals. d Experimental validation of miRNA-mediated edQTL:eQTL colocalization events. Diagram of the 3′-UTR reporter vector using plasmids containing 3′-UTR fragments with the edQTL RNA editing sites (top). Luciferase mRNA levels measured by qPCR of tested 3′-UTR constructs in the presence of editing-specific miRNAs (bottom left). Each barplot displays the mean and SD of three independent experiments. *p < 0.05, **p < 0.01. Diagram indicating the editing-specific targeting of miRNAs to unedited or edited 3′-UTR sequences (bottom right). Guanine was used in place of inosine to indicate an edited site