Abstracts Category : Other

Post-GWAS Bioinformatics and Functional Analysis ofDisease Susceptibility Loci

by Paul Martin

Genome-wide association studies (GWAS) have beentremendously successful in identifying genetic variants associatedwith complex diseases, such as rheumatoid arthritis (RA). However,the majority of these associations lie outside traditional proteincoding regions and do not necessarily represent the causal effect.Therefore, the challenges post-GWAS are to identify causalvariants, link them to target genes and explore the functionalmechanisms involved in disease. The aim of the work presented hereis to use high level bioinformatics to help address thesechallenges.There is now an increasing amount of experimental datagenerated by several large consortia with the aim of characterisingthe non-coding regions of the human genome, which has the abilityto refine and prioritise genetic associations. However, whilstbeing publicly available, manually mining and utilising it to fulleffect can be prohibitive. I developed an automated tool,ASSIMILATOR, which quickly and effectively facilitated the miningand rapid interpretation of this data, inferring the likelyfunctional consequence of variants and informing furtherinvestigation. This was used in a large extended GWAS in RA whichassessed the functional impact of associated variants at the 22q12locus, showing evidence that they could affect generegulation.Environmental factors, such as vitamin D, can alsoaffect gene regulation, increasing the risk of disease but aregenerally not incorporated into most GWAS. Vitamin D deficiency iscommon in RA and can regulate genes through vitamin D responseelements (VDREs). I interrogated a large, publicly available VDREChIP-Seq dataset using a permutation testing approach to test forVDRE enrichment in RA loci. This study was the first comprehensiveanalysis of VDREs and RA associated variants and showed that theyare enriched for VDREs, suggesting an involvement of vitamin D inRA.Indeed, evidence suggests that disease associated variantseffect gene regulation through enhancer elements. These can actover large distances through physical interactions. A newlydeveloped technique, Capture Hi-C, was used to identify regions ofthe genome which physically interact with associated variants forfour autoimmune diseases. This study showed the complex physicalinteractions between genetic elements, which could be mediated byregions associated with disease. This work is pivotal in fullycharacterising genetic associations and determining their effect ondisease. Further work has re-defined the 6q23 locus, a regionassociated with multiple diseases, resulting in a majorre-evaluation of the likely causal gene in RA from TNFAIP3 toIL20RA, a druggable target, illustrating the huge potential of thisresearch. Furthermore, it has been used to study the geneticassociations unique to multiple sclerosis in the same region,showing chromatin interactions which support previously implicatedgenes and identify novel candidates. This could help improve ourunderstanding and treatment of the disease.Bioinformatics isfundamental to fullyAdvisors/Committee Members: Eyre, Stephen.