Fig. 5

Distinct families of repetitive elements contribute to recently evolved and maintained regulatory regions. a Relative enrichment of recently evolved tissue-shared versus tissue-specific regulatory regions for selected transposable element families shown as a heatmap. Within each family, significance of tissue-specific vs. tissue-shared proportions calculated with the z-test and Bonferroni correction (p values ***< 0.001; **< 0.01; *< 0.05; − < 0.1. See Additional file 1: Figure S9B for maintained regions). b Validation of tissue-specific activity using the average ChIP-seq read enrichment for recently evolved active promoters associated with LINE L1s and L2s and their flanking regions. c Distribution of RNA-seq read counts for the promoter flanking regions in b. Dotted lines represent the median of tissue-specific RNA-seq enrichment for the tissue profiled. P values calculated using one sided the Wilcoxon test; within row test if read counts in each LINE-associated region type (column) is greater than in all other regions. d Estimated age of LINE L1s and L2s, as inferred by the number of substitutions from consensus sequence. LINE L1s that overlap regulatory regions (medium and light gray) are significantly older than inactive L1s (dark gray), while regulatorily active L2s are of similar age to inactive L2s. Dotted line is the median % divergence of the corresponding regulatorily inactive LINEs and p values calculated using one sided the Wilcoxon test for greater sequence divergence. Divergence is shown for all ten species combined; see Additional file 1: Figure S5B for per-species divergences. e Heatmap of relative enrichment in transposable element families for regulatory regions with evolutionarily dynamic (switch) versus stable signatures. Within each family, significance of evolutionarily dynamic vs. stable proportions calculated with the z-test and Bonferroni correction (p values ***< 0.001; **< 0.01; *< 0.05; − < 0.1)