WARNING: the interactive features of this website use CSS3, which your browser does not support. To use the full features of this website, please update your browser.
A study where machine learning was used to identify risk factors for incident heart failure. Leg bioimpedance was identified as a new marker which was further explored in confirmatory analyses.
Phenome-wide association studies of drug targets
Cardiovascular disease, including heart disease and stroke, is the global leading cause of morbidity and mortality. Recent advances in human genetics open up new avenues for investigation of downstream effects of perturbation of drug targets. By investigating the phenome-wide characteristics of individuals carrying gene-disrupting alleles, we can characterize the phenotypic effects as a function of the number of functional alleles of specific genes. Using this human knockout model, we can simulate and predict what the result would be if blocking the corresponding protein. Knowledge about associations of novel and known drug targets with cardiovascular outcomes as well as with other phenotypes will give important insights that can accelerate development of new treatment strategies. Hence, the proposed research does meet UK Biobank's stated purpose via improving the prevention and treatment of chronic disease. Specific Aim 1: To predict potential for repurposing or cardiovascular side effects of known drug targets using genetic information
We will perform single variant analyses, as well as burden tests (combining all protein-disrupting variants within a gene), for all genes encoding known drugs targets in relation to cardiovascular outcomes, as well as with common cardiovascular side effects.
Specific Aim 2: To predict phenome-wide effects of drugs developed against cardiovascular disease
We will perform association analyses of drugs currently or previously in development for treatment of cardiovascular disease with the whole range of phenotypes available in the UK Biobank. Full cohort