Large Scale Meta-analysis Characterizes Genetic Architecture for Common Psoriasis-associated Variants

The genotypes obtained from the Affymetrix Biobank Plus array constituted one of eight Caucasian datasets used in a large genome-wide meta-analysis of the association of psoriasis cases versus unaffected controls. With a combined effective sample size of > 39,000 individuals, this study identified 16 new psoriasis loci achieving genome-wide significance, increasing the total number of identified psoriasis loci for European-ancestry populations to 63.]]>
The purpose of this analysis was to test for genetic associations with psoriasis using a large case-control sample typed on the Affymetrix Biobank Plus array. Post-QC sample size is 11,674 people (5,960 psoriasis cases and 5,714 unaffected controls), all of European Caucasian ancestry. This association analysis was conducted solely for this submission to dbGaP; the primary reference study for these data (Tsoi et al., 2017, "Large scale meta-analysis characterizes genetic architecture for common psoriasis-associated variants.", Nature Communications) used only a portion of the samples typed on the Affymetrix Biobank Plus array as part of a large meta-analysis of 8 Caucasian datasets for psoriasis associations. The full post-QC dataset was used in this analysis. Separate analyses are being provided for two subsets of this sample set that were consented differently (GRU and DS). Please note that while assocation and other summary statistics are being made available for the entire sample set typed on this array platform, individual-level genotype data are being submitted for only 9,694 samples 5,476 cases and 4,218 controls), because controls collected at the University of Utah and all samples collected at Linköping University in Sweden were consented in a manner that does not allow sharing of their genotypes.
]]>Psoriasis cases were diagnosed by a dermatologist. Controls were at least 18 years of age and unaffected with psoriasis. All study participants were of self-reported European Caucasian ancestry. Cases and control individuals were recruited at five different locations - Michigan and Utah in the United States, Toronto and Newfoundland in Canada, and in Sweden. Samples with > 2% missing genotypes on the Affymetrix Biobank Plus array, an absolute value of the F inbreeding coefficient > 0.03, significant non-Caucasian admixture by principal components analysis, or that were duplicates of or highly related to (Plink Pi_hat > 0.25) individuals within this dataset or with the other seven datasets of Tsoi et al. (2017) were removed before analysis. After quality control, 7,891 individuals from the Affymetrix Biobank Plus dataset (3,863 cases and 4,028 controls) were used for the meta-analysis of Tsoi et al (2017). However, the dataset deposited here (5,942 cases and 5,732 controls) represents application of all quality control filters except the removal of samples that were duplicates or relative pairs with individuals of the other seven datasets. ]]>