Identifying causes of obesity and liver disease using population genetics

Obesity and nonalcoholic fatty liver disease (NAFLD) are global epidemics that cause much morbidity and mortality. We aim to use UK Biobank data to carry out genome wide association analyses on obesity and liver related traits to 1. identify and characterize genetic variants that associate with these traits 2. identify disease subtypes using cluster analysis of genetic variant effects across diverse metabolic diseases and traits 3. determine the causal relationship between a genetic variants and traits using mendelian randomization. These studies will identify causes of these diseases and improve diagnosis, management and care for people suffering from these conditions. Obesity and nonalcoholic fatty liver disease (NAFLD) are global public health epidemics. NAFLD is caused by fat deposition in liver and can progress to nonalcoholic steatohepatitis (NASH) and fibrosis/cirrhosis of the liver. NAFLD will become the number one cause of liver disease worldwide by 2020. Obesity increases risk of developing metabolic diseases such as diabetes and cardiovascular disease and cancers. There are few effective medical treatments for obesity and NAFLD in part due to a poor understanding of their etiology. Our work will elucidate the etiology of obesity/NAFLD and thus improve diagnosis, management and treatment of these diseases. We will identify true correlations between genetic variants and human traits using regression analysis where we quantitate the effect of these correlations controlling for variables that could artificially cause such correlations. We will examine the effects of associated variants on metabolic, neurological and cancer traits and then group these variants by their pattern of effects across these traits (cluster analyses) to identify subtypes of disease. Finally, we will model the genetic effects of variants on phenotypes using mendelian randomization to define the primary versus secondary effects of genetic variants. We will use the maximum number of individuals with available phenotypes and genotypes to improve power to find reproducible associations. We anticipate updating analyses as more samples and data become available.