Extended Data Fig. 7: Location of H1-hESC predictions.
From: Supervised enhancer prediction with epigenetic pattern recognition and targeted validation
a) The probability density of the distance of the predicted promoter and enhancer from the closest TSS is shown. b) The location of the enhancers and promoters on genomic elements are shown. Promoters are defined as TSS +/- 2kb. All TSS, UTR, exons, introns, and intergenic elements are calculated based on GENCODE 19 definitions. A regulatory region is considered to overlap with the elements if more than 50% of the matched filter region overlaps with the corresponding element in b. c) The distribution of gene expression of gene closest to the enhancer/promoters are plotted and compared to the gene expression of all genes in H1-hESC. A two-sided Wilcoxon test shows that P-value for differences in gene expression of genes close to enhancers and promoters are significantly higher than expression of all genes in H1-hESC (< 10−100 each). The center line in each category represents the median expression level for all genes close to corresponding category while the lower and upper boundaries of the box indicate the 25th and 75th percentile of the expression levels for genes within that category.
