New Computational Tools for Advanced Analytics in Genome-wide Association Studies

点击翻译按钮获取中文摘要

Project Summary Many genome-wide association studies (GWAS) have been successfully carried out, identifying numerous genetic variants associated with common diseases and disease related complex traits. The identified genetic associations can now explain a large fraction of the genetic contribution and trait heritability, revealing the genetic architecture underlying common diseases. The success of GWAS in the past few years have laid down a solid foundation both for pursuing the mechanistic insights towards the biology of disease and for transitioning towards new diagnostics and therapeutics in clinical settings. Advancing GWAS towards mechanistic insights and clinical translations, however, urgently requires the development of advanced computational methods that can take advantage of the unique data features and increased data complexity of GWAS as well as the data available from parallel genomics studies. Here, we propose to develop a set of new computational methods to advance GWAS analytics beyond simple variant association analysis and move towards the understanding of the biology of disease and enable potential clinical translations. Specifically, in Aim 1, we will develop integrative methods to integrate GWAS with multiple gene expression mapping studies of distinct genetic ancestries to investigate the molecular mechanisms underlying the variant-trait associations and interrogate the contribution of ancestry specific genetic architecture underlying expression towards gene-trait associations. In Aim 2, we will develop causal inference methods to leverage the genetic associations to improve our understanding of the causal relationship among complex traits and to identify causal risk factors that underlie disease etiology. In Aim 3, we will develop prediction methods to make use of the genetic associations and take advantage of the genetic and environmental correlation among multiple complex traits to facilitate the genetic prediction of disease risk, aiding disease diagnosis and clinical applications. All methods will be implemented in user-friendly open- source software and disseminated to the scientific community. At its conclusion, the proposed study will provide a comprehensive suite of computational methods and software tools for advanced analytics in GWAS. These methods are essential for understanding the transcriptomic and causal mechanism underlying disease etiology, enabling accurate and robust genetic prediction of disease risks, and facilitating biological discoveries and insights.