The University of Manchester, Ashley Budu-Aggrey, PhD, Investigation of a susceptibility locus for Psoriatic Arthritis, 2016Psoriatic arthritis (PsA) is an inflammatory arthritis that is associated with psoriasis. PsA is a complex disease that is influenced by both genetic and environmental factors. Genetic studies have been successful in identifying risk loci for PsA, the majority of which also confer risk for psoriasis. As PsA has been estimated to have a stronger genetic component compared to psoriasis, this suggests that there should be genetic loci that are associated with PsA and not with psoriasis, which I have defined to be PsA-specific. Therefore, the overall aim of my PhD project is to apply genetic experimental techniques to identify a novel risk locus that is associated with PsA and not psoriasis.Firstly, genotyping of 1,962 PsA cases was performed with the Immunochip, where I contributed to the identification of a novel PsA-specific association on chromosome 5q31, mapping to rs715285 (P=4.38x10-13). Bioinformatic analysis within the association region identified 3 potential causal genes and 4 potential causal variants. Statistical analysis identified CD8+ memory T cells to be the most critical cell type for PsA, and also gave evidence to suggest that rs10065787 is the causal variant within the 5q31 region. A cell-specific eQTL study was performed within this cell type which identified SLC22A5 to be a potential causal gene for PsA. The findings may suggest a potential mechanism by which the 5q31 region contributes towards PsA susceptibility. Variants that reached nominal association during the Immunochip study (P <1.0x10-4) were selected to follow-up in a candidate SNP study, in addition to other SNPs of interest which had been investigated for psoriasis risk but not PsA alone. This was performed within an independent cohort (914 PsA cases, 6,945 controls). Analysis of the independent samples replicated a PsA-specific association with rs12044149 (P = 4.03x10-6) at the IL23R locus. This association reached genome-wide significance during meta-analysis with the discovery Immunochip genotype data (P = 1.28x10-20). The PsA association with IL23R was found to be independent of the psoriasis association at the same locus when performing conditional analysis (Pcond = 4.86x10-06). The effect estimates for rs12044149 in PsA and psoriasis were found to be significantly different (P = 2.0x10-3). Furthermore, direct comparison of the genotypes for rs12044149 showed a significantly increased risk for PsA compared to psoriasis. Functional annotation was performed with credible SNPs sets defined for the PsA and psoriasis association signals at IL23R. This also revealed potential differences in the mechanisms of the two diseases, supporting the independence of their association with IL23R.A rare variant study was also performed as part of this PhD in order to identify a casual gene for PsA. 1,980 PsA cases and 5,913 controls were genotyped with the Illumina HumanCoreExome and HumanExome chips. Single-variant analysis was performed using the Fisher's Exact test. Analysis of common variants (MAF > 0.05) uncovered associations at known PsA and psoriasis loci that reached genome-wide significance, including TRAF3IP2 (P = 3.88x10-18). Analysis of rare (MAF <0.01) and low-frequency (MAF <0.05) variants identified an association with a coding-variant at IFIH1 (rs35667974, P = 2.39x10-6). This association was replicated within an independent cohort, and was found to be significantly associated during multiple-variant testing (P = 2.30x10-6, Pcorr = 5.60x10-6). The minor allele of the rare variant was found to exist in the same haplotype as the minor allele of the PsA-associated common variant at the IFIH1 locus. The rs36557974 SNP found to be associated with PsA independently of the common variant during conditional logistic regression and conditional haplotype analysis. Although the rare-coding variant has been previously reported for psoriasis, these findings demonstrate the use of rare variant analysis to find evidence of a casual gene for PsA.In summary, genetic and statistical approaches have been used to find evidence of causal genes and variants within PsA-specific association regions, as well as to identify a PsA association mapping to a coding region.