Figure 3. Motif analysis reveals miRNA binding dependent on seed and auxiliary pairing.

The proportion of chimera-defined target regions with the indicated seed variants is plotted, broken down (a) by transcript region, (b) by the number of times interactions were identified with chimeras (N) or whether chimeras overlapped AGO CLIP peaks and (c) for the most abundant miRNA families in mouse brain, ranked from the top by decreasing abundance. (d) Overlap of 3′-UTR chimera-identified sites in the brain with TargetScan predicted sites for the same miRNA (red) or three equally sized random control sets of TargetScan sites (black). Control sets were restricted to the top 20 brain miRNAs. Only target sites in mRNAs with detectable expression in the cortex were considered. (e) The distributions of mismatched and bulged nucleotides for chimera-identified sites with imperfect seed motifs are plotted for the top 25 mouse brain miRNAs (black), miR-124 (red) and miR-9 (blue). Error bars show the s.d. at each position for the top 25 miRNAs in the brain. miRNA seed sequences for miR-124 and miR-9 are shown below mismatch and miRNA bulge plots. Below the target bulge plot, the most frequently bulged target nucleotide at the indicated position is shown when strong preferences (>50% of sites) were apparent. Sites from all transcript regions were included in this analysis. (f) De novo analysis of cognate miRNA-complementary-enriched 7mer motifs in all chimera target regions plotted as a heat map across the miRNA. Each line represents one miRNA and colour intensity scales with abundance in target sequences. miRNAs are ordered by hierarchical clustering.