Fig. 8: Allele specific binding of epidermal homeostatic TFs.

a Experimental diagram for identifying ASB from CUT&RUN data. b Number of SNVs in each category of ASB. The number of reads from each allele in a heterozygous site are tested by a two-sided binomial assuming equal probability of binding between alleles. FDR was estimated by Benjamini-Hochberg. c Numbers of putative ASBs by indicated ASB category for TFs studied. d Concordance of predicted motif strength change and the direction of altered binding; note good agreement for sequence-specific enhancer-binding proteins, but lower agreement for SP/KLF factors that mostly associate with common C-rich sequences in promoters. e Absolute log2 fold change in reads at ASBs as a function of genomic location. Whiskers, 10–90% data range. P value, two-sided Mann–Whitney. Center line, median. Bounds of box, 25–75% data range. Outliers are not drawn due to the high number of datapoints. “N” is the number of bound heterozygous sites. f Bar height is the fraction of heterozygous SNV-protein pairs with at least 20 CUT&RUN reads that are putative (P < 0.1) ASBs as a function of genomic location. P value, two-sided Mann-Whitney. Whiskers, 95% CI. “N” is the number of bound heterozygous sites. g Putative ASBs are enriched near skin disease genes relative to expressed genes (P value test for greater enrichment vs random expressed gene selection was determined by taking random subsets of expressed genes and determining what fraction of random selections exceeded the observed number). h 90% of ASBs in skin disease-linked SNVs are eQTLs. P value, two-sided Fisher’s exact. i Number of ASB events near epithelial homeostasis genes in the catalog.