Supplementary Figure 4: Additional data related to eQTL - SuRE comparisons.
From: High-throughput identification of human SNPs affecting regulatory element activity
a. Odds ratios of concordance of whole blood eQTL and K562 raQTL SNPs, as a function of maximum distance to the TSS of the associated eGene. All odds ratios are significantly larger than 1 (one-sided Fisher’s exact test, P<3.6e-3 for all distance cutoffs). Analysis based on 623,210 SNPs that overlap between SuRE and eQTL datasets. b. Same as (a) but for liver eQTL SNPs that are raQTLs in HepG2. All odds ratios are significantly larger than 1 (P<2.3e-4 for all distance cutoffs). Analysis based on 186,613 SNPs that overlap between SuRE and eQTL datasets. c. Genome track plot combining SuRE data and eQTL mapping data for LYZ in whole blood, similar to main Fig. 4e. d. Protein binding analysis for rs554591, similar to main Fig. 4f. e. Barplots indicating fraction of reads containing each of the two alleles for rs623853 in K562 genomic DNA (left) and K562 DNase-seq reads (Encode Project Consortium, 2012, Nature 489, 57-74) (right). P-values are from one-sided Fisher exact test. f. Same as (e) but for rs554591.
