Celiac.com 03/11/2010 - As part of an effort to investigate the possibility of multiple common variants for celiac disease influencing immune gene expression, a team of more than sixty scientists recently worked together to conduct a second-generation genome-wide association study (GWAS) of 4,533 individuals with clinically proven celiac disease, along with 10,750 control subjects.
The GWAS included five European sample collections of celiac disease and control cases, including the celiac disease dataset reported previously. The team's stringent data quality control measures included calling genotypes using a custom algorithm on both large sample sets and, where possible, cases and controls together. The team tested 292,387 non-HLA SNPs from the Illumina Hap300 marker pool for association in 4,533 individuals with celiac disease and 10,750 control subjects of European descent. They also tested another 231,362 additional non-HLA markers from the Illumina Hap550 marker set for association in a subset of 3,796 individuals with celiac disease and 8,154 controls. All markers came from autosomes or the X chromosome. For both datasets, Genotype call rates were >99.9%.
The study showed over-dispersion factor of association test statistics comparable to that observed in other GWASs of this sample size. Factoring in missing genotypes for 737 cases with celiac disease genotyped on the Hap300 BeadChip and corresponding controls did not change the findings in any meaningful way.Variants from 13 new regions reached genome-wide significance (Pcombined < 5 Ã— 10âˆ’8); most contain geneswith immune functions, such as BACH2, CCR4, CD80, CIITA-SOCS1-CLEC16A, ICOSLG and ZMIZ1, while ETS1, RUNX3, THEMIS and TNFRSF14 play key roles in thymic T-cell selection.
The data suggested associations for 13 additional regions. Expression quantitative trait meta-analysis of 1,469 whole blood samples showed that 52.6% of tested loci (20 of 38 loci) had celiac risk variants corresponding with cis gene expression (P < 0.0028, FDR 5%).