The development of new and effective therapies for type 2 diabetes (T2D) requires the identification of novel drug targets, ideally ones that are validated by strong evidence of clinical benefit from studies in human populations. Inherited loss of function (LoF) variants offer one approach to assess the impact of reducing gene activity in humans in vivo. Particularly strong evidence for target validation can be obtained by observing LoF variants that provide protection against disease without undesirable consequences (as in CCR5 and PCSK9). The applicants have collected high coverage exome sequencing data in DNA samples from each of 2,800 individuals (T2D cases and controls), and will have genotyped a comprehensive collection of non-synonymous protein altering variants in >45,000 individuals (T2D cases and controls) using the "exome" array. To perform a systematic and well powered analysis of these data for LoF variants, several challenges must be overcome: (a) development and application of algorithms for robust detection of insertion and deletion variants (a major mechanism for LoF variants which is poorly characterized with today's algorithms), and for the accurate annotation of all classes of LoF variants;(b) characterization of statistical tests that re sensitive for the frequency spectrum and characteristics of LoF variants, and their application to the catalogue of LoF variants detected in cases and controls;and (c) follow-up of putative LoF associations in large, independent samples. To test systematically the role of rare protein-altering LoF variants in risk of T2D, the applicants propose: (a) to develop algorithms to detect indels in sequence data, and larger deletions using data from the exome array;to apply these algorithms to sequence and genotype data totalling >47,000 DNA samples;and to annotate LoF variants across the genome;(b) to evaluate the power of rare variant tests for LoF analysis, and to perform association analyses using chosen methods for LoF variants, both singly and for sets of LoF variants within a gene, with T2D, as well as with the related metabolic traits of glucose, insulin, lipids, and BMI;and (c) to validate putative associations of LoF variants with altered rik of T2D (in particular, protection from T2D) by performing in silico follow-up in data on up to 10,000 individuals (T2D cases and controls) from the T2D-GENES Project, and by targeted sequencing in 20,000 additional individuals (T2D cases and controls). A central goal of human genetics research is to provide insights that can guide breakthrough approaches to prevention and therapy. The applicants have been leaders in the development of datasets, laboratory methods, and algorithms for genetic analysis, and have collaborated for over a decade to apply these methods to discover genes for T2D. Now, the convergence of large clinical samples from T2D cases and controls, of next-generation sequencing technology, and of algorithmic improvements make it possible to evaluate systematically LoF variants for effects on T2D, nominating and validating potential therapeutic targets.
The rising prevalence of type 2 diabetes in the US and worldwide represents one of the major challenges to public health, and improved options for treatment and prevention are required. The present proposal builds on a longstanding and productive collaboration between researchers in the US and Europe to understand the genetic basis of type 2 diabetes, and to use this information to reveal disease mechanisms. In this proposal, we will focus on the subset of DNA sequence variants that have the most dramatic effects on gene function (loss of function coding variants), and seek to define the role that these play in type 2 diabetes predisposition.
|Rivas, Manuel A; Pirinen, Matti; Neville, Matthew J et al. (2013) Assessing association between protein truncating variants and quantitative traits. Bioinformatics 29:2419-26|
|Ma, Clement; Blackwell, Tom; Boehnke, Michael et al. (2013) Recommended joint and meta-analysis strategies for case-control association testing of single low-count variants. Genet Epidemiol 37:539-50|