Dissection of a Complex Disease Susceptibility Region Using a Bayesian Stochastic Search Approach to Fine Mapping

Abstract

Identification of candidate causal variants in regions associated with risk of common diseases is complicated by linkage disequilibrium (LD) and multiple association signals. Nonetheless, accurate maps of these variants are needed, both to fully exploit detailed cell specific chromatin annotation data to highlight disease causal mechanisms and cells, and for design of the functional studies that will ultimately be required to confirm causal mechanisms. We adapted a Bayesian evolutionary stochastic search algorithm to the fine mapping problem, and demonstrated its improved performance over conventional stepwise and regularised regression through simulation studies. We then applied it to fine map the established multiple sclerosis (MS) and type 1 diabetes (T1D) associations in the IL-2RA (CD25) gene region. For T1D, both stepwise and stochastic search approaches identified four T1D association signals, with the major effect tagged by the single nucleotide polymorphism, rs12722496. In contrast, for MS, the stochastic search found two distinct competing models: a single candidate causal variant, tagged by rs2104286 and reported previously using stepwise analysis; and a more complex model with two association signals, one of which was tagged by the major T1D associated rs12722496 and the other by rs56382813. There is low to moderate LD between rs2104286 and both rs12722496 and rs56382813 (r2 ≃ 0:3) and our two SNP model could not be recovered through a forward stepwise search after conditioning on rs2104286. Both signals in the two variant model for MS affect CD25 expression on distinct subpopulations of CD4+ T cells, which are key cells in the autoimmune process. The results support a shared causal variant for T1D and MS. Our study illustrates the benefit of using a purposely designed model search strategy for fine mapping and the advantage of combining disease and protein expression data.

Chromosome contacts in activated T cells identify autoimmune disease candidate genes

“…Tables S6a-S6b, and show that of 73 regions with genetic association signals to at least one disease (minimum p < 5x10 -8 ; 106 disease associations), nine regions have strong evidence that they contain more than one causal variant (posterior probability > 0.5), among them the well studied region on chromosome 10 containing the candidate gene IL2RA [25] . For the GWAS summary data, we make the simplifying assumption that there exists a single causal variant in any LD-defined genetic region and again generate posterior probabilities that each variant is causal [26] .…”

“…Multiple variants in and near IL2RA have been associated with a number of autoimmune diseases [34,[48][49][50] . We have previously fine-mapped genetic causal variants for T1D and multiple sclerosis (MS) in the IL2RA region [25] , identifying five groups of SNPs in intron 1 and upstream of IL2RA, each of which is likely to contain a single disease causal variant. Out of the group of eight SNPs previously denoted "A" [25] , three (rs12722508, rs7909519 and rs61839660) are located in an area of active chromatin in intron 1, within a PIR that interacts with the IL2RA promoter in both activated and non-activated CD4 + T cells ( Fig.…”

Accuracy of Gene Scores when Pruning Markers by Linkage Disequilibrium

“…On the other hand, if there are several causal variants in mutual LD, then it may be preferable to include all those variants -but only those -in the score. Identifying the casual variants within a region of LD is an important problem when describing aetiology, and there is an extensive literature of statistical methods for this purpose [7][8][9][10] . For epidemiological applications such as risk prediction or patient stratification, however, the aim is often to derive a parsimonious but accurate model that is not necessarily aetiological.…”

Improving the coverage of credible sets in Bayesian genetic fine-mapping

“…Early statistical approaches for fine-mapping tended to focus on the SNP in the region with the smallest P 11 value, called the lead-SNP. However, it is generally acknowledged that this SNP may not be the causal variant 12 in a given region due to correlations with the true causal variants [1,4]. Studies may therefore prioritise the 13 lead-SNP before extending the analysis to include either variants in high LD with this SNP or the top k 14 variants with the highest evidence of association [5].…”

“…Our method is limited in that it does not model multiple causal variants. Fine-mapping approaches that 181 jointly model SNPs have been developed, such as GUESSFM [4] which uses genotype data and FINEMAP [11] 182 and JAM [14] which attempt to reconstruct multivariate SNP associations from univariate GWAS summary 183 statistics, differing both in the form they use for the likelihood and the method used to stochastically search 184 the model space. The output from these methods are posterior probabilities for various configurations of 185 causal variants, and therefore the grouping of SNPs to distinct association signals must be performed post-hoc 186 to obtain similar inferences to that of single causal variant fine-mapping (e.g.…”