Fig. 1
Benchmarking the ImmunoSeq capture panel by known disease associated sites and regulatory variants. a Autosomal GWAS hits associated to more than one autoimmune or chronic inflammatory disease, for neuropsychiatric diseases and for cancer included in the Immunoseq. custom capture panel. (Cut-off of 1 × 10−8 was used to select GWAS hits to analyze, SNPs in LD selected based on r 2 > 0.9, HLA (human leucocyte antigen) hits and region as well as chromosome X SNPs were excluded from the analyses). SNP in LD = GWAS hits that have a SNP in LD in the Immunoseq. custom capture panel. b cis-eQTLs from monocytes (CD14+) and B Cells (CD19+) (considered has haplotype block, r 2 > 0.9) included in the Immunoseq. panel. Cut-off of p < 1e-3 or p < 1e-5, and p < 1e-12 after 1000 permutations (1000 = number of SNPs tested per probe) and top 1 eQTLs per transcript were kept for analysis (HLA hits and region as well as chromosome X hits were excluded in the analyses). c Enrichment of GWAS hits (same as in A) and proximal SNPs (LD r 2 > 0.9) that fall in DHSs selected for immune cell types compared to DHSs selected from other tissues (either all or non-overlapping ones) and regions randomly selected (1000 times) from the whole genome (either the full genome or only non-coding regions excluding HLA). Significance was calculated using Fisher’s exact test. Enrichment is significant (p < 0.001) for all GWAS hits except for Neuropsychiatric hits. d Enrichment of eQTLs (same as in B) and proximal SNPs (LD r 2 > 0.9) positioned at DHSs selected for immune cell types compared to DHSs selected from other tissues (either all or non-overlapping ones) and regions randomly selected (1000 times) from the whole genome (either entire genome or only the non-coding part excluding the HLA region). All enrichments shown are significant (p < 0.001). All p-values were calculated using Fisher’s exact test