Fig. 3

a Representative subset of differentially methylated 200-bp tiles across the 26 DyMe-Seq datasets (N = 50,000, randomly selected) generated from the entire GTEx pilot cohort (n = 25). Skin nse (not sun exposed) 1 and 2 are technical DyMe-Seq replicates from the same DNA samples. Thyroid 1–3, skin 1–3 and lung 1–3 are samples from three different individuals. Each row represents one 200-bp tile, and the color indicates the methylation level (black unmethylated, red highly methylated). b Fraction of differentially methylated (dynamic) and static 200-bp tiles across the GTEx cohort. A tile was called differentially methylated if the methylation difference exceeded 0.3 at a q value ≤0.05 (Fisher’s exact test, BH-corrected) between any of the samples. c Overview of number of DMRs across all pairwise comparisons within the GTEx cohort. Here, differentially methylated tiles within 400 bp of each other were merged into DMRs. d Gene set enrichment analysis for comparatively hypomethylated regions between the GTEx sample coronary aorta and nerve tibial. Shown are selected gene set categories from the top 40 significantly enriched gene sets (see “Methods” section for details). e Top 15 transcription factor motifs enriched in regions differentially methylated between coronary aorta and nerve tibial based on the transcription factor epigenetic remodeling activity (TERA) framework using the a de-methylation score. Transcription factors with high scores are likely to bind frequently to comparatively hypomethylated regions in the respective cell type