Towards an integrated map of causal connections for common, complex diseases

点击翻译按钮获取中文摘要

PROJECT SUMMARY / ABSTRACT Research in my laboratory aims to understand the genetic and biological links between different complex traits and diseases. We plan to use computational approaches from several different disciplines, including human genetics, statistical genetics, epigenetics and population genetics to map out the genetic, functional and evolutionary links between hundreds of traits and diseases simultaneously. Genetic studies have inferred causal connections among numerous traits (re. measurable indicators of the severity or presence of a disease state) and disease. However, almost all of these studies have tested one trait with one disease at a time. While important for testing specific hypotheses about specific relationships obtained from epidemiological studies, these studies, by nature, tend to miss unforeseen and unexpected causal connections with other traits or disease. Furthermore, complex patterns across several traits and diseases would be missed. For these reasons, there is a need to consider an approach that incorporates all traits and disease links in a single, unified framework (the `phenome-wide map'). To this end, over the next five years, we plan to embark on a series of studies to first, 1) build a phenome- wide map of causal connections between a multitude of common human traits and diseases. This map requires individual single nucleotide variant (SNV) level association results from genome-wide association data. As a result, I have begun to build a comprehensive repository of genome-wide association data for millions of SNVs and hundreds of different traits, biomarkers and diseases from several studies. This data spans a wide spectrum of common diseases, including cardiovascular disease, cardiometabolic conditions, inflammatory diseases, psychiatric disorders, renal function, amongst others. We will infer causal connections using this repository of data between all combinations of associated SNVs, traits and diseases to generate the phenome-wide map. Next, we will add biological links to the map by incorporating information related to 2) molecular function via gene regulation. We will infer links to each SNV in the phenome-wide map with regulatory elements, cell types, and expression of genes. Third, we will incorporate 3) natural selection metrics at the per gene level into our phenome-wide map. We will develop an approach to make predictions on the strength and mode of natural selection at the per gene level, and then add this to our map of causal connections. Finally, we expect to use the phenome-wide map to explore similarities and differences across the different links observed between the traits and diseases. Our proposed research program can provide insights into new biological mechanisms behind the shared etiology of traits and diseases. Importantly, our research also has direction precision medicine applications as it can inform about prioritization of new gene targets for drug discovery efforts.