Table 1. Number of cis-eQTLs driven by the same causal variant as the SLE disease association (total number of SLE loci: 38).
Gene | Transcript | Exon | Junction | Intron | Total | |
---|---|---|---|---|---|---|
Causal cis-eQTLsa | 2 | 2 | 7 | 4 | 4 | 9b |
% of 38 SLE GWAS loci | 5.3 | 5.3 | 18.4 | 10.5 | 10.5 | 23.7 |
% of total causal eQTLs | 22.2 | 22.2 | 77.8 | 44.4 | 44.4 | 100 |
Candidate genes | 3 | 4 | 9 | 5 | 5 | 12 |
Expression targetsc | 2 | 7 | 24 | 18 | 13 | 64 |
The lead SNPs from the Bentham and Morris et al 2015 GWAS in persons of European descent were functionally annotated by cis-eQTL analysis in the Geuvadis RNA-Seq cohort in lymphoblastoid cell lines using RNA-Seq quantification profiled at five resolutions (gene, transcript, exon, junction, and intron). Only SNPs reaching genome-wide significance, not conditional peaks, outside of the major histocompatibility complex loci, and with minor allele frequency > 5% were included leaving 38 SLE lead SNPs in total. All SLE loci were densely imputed to the 1000 Genomes Phase 3 Imputation Panel as described in methods.
All 38 loci (+/-100kb of each lead SNP) comprised a nominally significant cis-eQTL (P<0.01) for at least one gene within +/-500kb of the lead SNP at each resolution of RNA-Seq. Evidence of a single shared causal variant at each locus was assessed using the Joint Likelihood Mapping (JLIM) algorithm as described in methods.
aNumber of loci where the disease association is consistent with a single shared effect for at least one cis-eQTL (P<0.01 and JLIM FDR adjusted P<0.05).
bThe total number of unique causal cis-eQTLs across all RNA-Seq quantification types.
cExpression targets corresponds to the quantification type in hand (i.e. number of exons at exon-level).