Fig. 5: TF-binding preference for the ALT allele.

a TF-binding density across the AXIN2 LCL VCM locus from GM12878 ENCODE data. b Percentage of reads associated to the ALT or REF-rs143348853 allele for significantly imbalanced TFs on the indel region (ATAC-seq peak of 755 bp overlapping the indel). TFs are sorted from left (low) to right (high) according to the FDR corrected p-value from a binomial test. TFs with a detected peak overlapping the region are labelled in bold. The percentage of ChIP-seq input reads is also shown. c Percentage of ALT reads from significantly imbalanced TFs for each region of interest: indel (n = 38), inside LCL AXIN2 VCM (n = 38), outside LCL AXIN2 VCM (n = 92) but same TAD, and the two adjacent TADs (n = 108), where n indicates significantly imbalanced TFs. Boxes indicate the IQR (25–75%) and the box center indicates the median. Whiskers represent de minimum or maximum values of no further than 1.5 times the IQR for both the top and bottom of the box. d The percentage of ALT reads from significantly imbalanced TFs for each CRE are plotted across particular regions (left panel) as well as the different, principal CREs (right panel) (CREs embedded in the LCL AXIN2 VCM). The dendrogram clustering based on Euclidean distances was performed only on the overall regions, and the TF list on the right was sorted accordingly. The promoter region was considered as the overlap of H3K4me3 and H3K27ac peaks. Enhancer regions are listed in Supplementary Table 1. To increase statistical power, the right-enhancer region was taken as the full H3K4me1 peak and not only the overlap with the H3K27ac peak. e Same as d but with all available histone marks. In c, d and e, the percentage of ALT associated reads is displayed as the log2 fold change of ALT read percentage of a TF or histone mark over the ALT read percentage of the input.