Skip to main content
Fig. 1 | Genome Biology

Fig. 1

From: Human A-to-I RNA editing SNP loci are enriched in GWAS signals for autoimmune diseases and under balancing selection

Fig. 1

The identification and verification of non-SNP and SNP editing sites. a The proportion of variants that are either A-to-G or T-to-C mismatches for SNP RNA variants in each sample. The GTEx and Geuvadis datasets were plotted, separately. All samples used for editing identification were shown. b Allelic ratio distributions for non-SNP editing sites, SNP editing sites, and control SNPs. SNP and non-SNP editing sites called in all individuals were plotted. Control SNPs are all known SNPs. For each control SNP, the allelic ratios for individuals with heterozygous genotype were calculated. Only samples with sites that are covered by at least 20 reads were used. c The nucleotides neighboring both the non-SNP and SNP editing sites show a pattern consistent with known ADAR preference. The motif is characterized by the underrepresentation of G upstream to the editing site. d Comparison of the hybridization energies between SNP editing regions (SNP editing sites and flanking ± 15 nt) and their predicted complementary sequences with those between shuffled editing regions and their predicted complementary sequences (see the “Methods” section). For each SNP editing site, we shuffled the editing region and predicted its complementary sequence. We repeated this 10,000 times and the mean value was calculated. The p value was calculated with the pairwise Wilcoxon rank sum test. e Comparison of the BLAST scores between SNP editing regions (SNP editing sites and flanking ± 25 nt) and shuffled editing regions. BLAST score represents the overall quality of an alignment (aligning the editing region to the genomic sequence ± 2000 nt of the SNP editing site, see the “Methods” section). The p value was calculated with the pairwise Wilcoxon rank sum test. f Boxplots showing the editing level changes of SNP editing sites and non-SNP editing sites upon ADAR1 or ADAR2 overexpression in HEK293 cells. g Boxplots showing the editing level changes of SNP editing sites and non-SNP editing sites between wild-type and ADAR1 knockout HEK293 cells. h The overlaps of SNP editing sites between GTEx and Geuvadis datasets

Back to article page