The NHLBI Exome Sequencing Project: Emerging Opportunities for Nephrology audioicon

While evidence for the contribution of rare, putatively functional, variants contributing to variation in disease risk and intermediate phenotypes is accumulating, there is also a need to better delineate the pathways and networks that can be resolved from existing genome-wide association scan (GWAS) and fine-mapping data. The Type 1Diabetes Genetics Consortium {T1DGC) conducted a GWAS meta-analysis and, with its increased sample size, identified over 40 loci that were significantly associated with T10 risk. However, each locus contained an average 7 genes in a region of 250 kb and, within each locus, there were some “obvious” candidate genes or only a single gene.It should be noted, however, that the “candidate gene” in each locus was often defined by the presence of a gene with known expression, immune function, association with other immune-mediated diseases,or other available genetic information. Specifically, the Mgene of interesf in each locus did not imply that this was the causalgene but reflected a bias of choice based upon anticipated effect on the immune system.

In order to refine GWAS regions osusceptibility for T10 and discover causal genes and risk variants, dense genotyping (fine mapping) was required, which led to the establishment of the lmmunoChip consortium {a consortium of investigators in nine autoimmune diseases who provided data, induding  unpublished, to design a genotyping chip for dense mapping). For T1D, we genotyped 6,671 UKGrid cases, 5,128 Birth Cohort controls, 2,277 ASP families and 1,074 trio families; a separate collection of 1,395 Birth Cohort controls and 2,893 UKBS controls were genotyped at the Wellcome Trust/Sanger Centre. In these data, we have discovered novelgenes for T10 in loci not included in the originalT1DGC GWAS meta­ analysis. These data from lmmunoChip, two sequencing and a follow-up genotyping study all point to the fact that some,If not most, genes associated with T1D risk may contain multiple common and rare variants. The implication of these results is that functional studies should not focus on the SNP (there will be many SNPs, common and rare, with uncertain effects),but on genes and pathways. An initial analysis using the causalgenes for T1D will be presented in a framework of T1D-causal pathways that can be used to generate more complete models of risk, pathogenesis and targets for perturbation. The results illustrate at least four important clusters. This Initial analysis provides an example of how construction of a series of prediction models can be implemented using only the genetic data from the lmmunoChip. Future research combining functionaldata in specific cell types (e.g., C04+ T cells), combined with common variant, rare variant, and other data, should greatly increase our understanding of the complex inheritance of T1D and associated phenotypes (diabetic nephropathy).

View the presentation below:

This was presented at the ISN Forefronts Symposium event “Systems Biology and the Kidney” that took place from 7-10 June 2012 in AnnArbor , Michigan, US.