A meta-analysis of 3 genome-wide association studies (GWAS) and 2 independant datasets genetopyed on the Immunochip using a total of 10,588 individuals with Psoriasis and 22,806 controls. Overall the anlysis included 111,236 SNPs that were genotyped in both Immunochip datasets and had good imputation quality in at least 2 of the 3 GWAS. The meta-analysis confirmed genome-wide associations (p < 5x10-8) at 19 of the 21 known loci and identified 15 new significant risk loci, bringing the total number to 36 in European individuals.
Pairwise comparison were perfromed using the genome function in PLINK to identify duplicate pairs or highly related individuals among the data sets. Principal component analysis was able to seperate samples by country of origin therefore a a linear mixed-model method (EMMAX) was used for the association analysis.
Imputation on cases and controls was performed using the CEU reference haplotypes to increase the number of overlapping SNPs between the datasets. SNPs with a low imputation quality (r2 ≤ 0.3 for minimac and info score < 0.5 for IMPUTE2) were removed.
Samples were excluded due to :-
- Low call rate(<98%).
- Non-European ancestry
- Exterme inbreeding co-efficients or heterozygosity values computed by PLINK
- Related sample pairs, most due to duplication
Markers were excluded due to :-
- Low call rate (<95%).
- Deviation from Hardy-Weinberg equilibrium in controls (p<1x10-06)
GDXHsS00016 meets the following criteria: