Heritability enrichment analysis, prediction and genetic correlation analysis for multiple traits.
Principal Investigator: Professor David Balding
Approved Research ID: 45918
Approval date: March 12th 2019
Many human traits related to health have a strong genetic component, but it has become increasingly apparent that the individual genetic contributions are often more numerous and more complex than previously imagined. Therefore, rather than study genetic contributions one at a time, we need to develop mathematical models to study simultaneously all the genetic influences on a trait of interest. We aim to do this and to go further: modeling all the genetic influences on many inter-related traits, all at the same time. This is extremely challenging because the human genome is vast and simple approaches to fitting a mathematical model will fail due to what statisticians call 'over-fitting'. There are many strategies for dealing with this problem, none of them perfect. We have already made considerable progress based on careful analysis of the inter-dependencies of genetic variants, by mining many large datasets from previous human genetics studies. Because of its large scale and high-quality data, UK Biobank provides an unprecedented opportunity to improve our model fitting and apply it to understand in fine detail which parts of the genome play what roles in different human traits. The next step will be to use our improved understanding of genetic architectures for better prediction of outcomes such as disease progression. We will investigate further downstream benefits from our very detailed genome-wide model of effects on multiple traits. Our research will lead to better understanding of genetic mechanisms underlying health and disease, and better prediction of outcomes. There is strong public interest in understanding in detail how our bodies work, a question that has long fascinated our ancestors but is now more than ever susceptible to precise answers. Better understanding will allow us to design better strategies for health maintenance and therapies for disease. Increased understanding may also improve public acceptance of unusual conditions. Similarly, while some people may not wish to predict their future health status, there are potential benefits for those who choose to, potentially including individually-tailored preventative health programs and targeted interventions, including novel pharmaceuticals.