Computational approaches to advance genomic, biological and clinical understandings of human disease

点击翻译按钮获取中文摘要

PROJECT SUMMARY / ABSTRACT The NIGMS Established Investigator (EI) R35 Maximizing Investigators’ Research Award (MIRA) proposal aims to use computational approaches to advance genomic, biological, and clinical understandings of human disease. The research program is broadly focused on three areas of human genomics: genomic association, biological mechanism and translational medicine. Within these areas, research in the laboratory has focused on: 1) evaluation of disease risk of genetic variants; 2) development and application of Mendelian randomization to infer causal relationships between complex traits and diseases; 3) evaluation of the complex interplay between natural selection and human diseases; and 4) using human genomics to inform drug side effect prediction. The proposed research program leverages large-scale genetic and clinical data resources, combined with statistical methods development, building directly on our prior published research in each of the research areas. Importantly, we have highlighted critical unmet needs, key knowledge gaps in our understanding and important challenges to be addressed pertaining to general medical sciences research. Over the next five years, we plan to embark on a series of studies designed to address these unmet needs and overcome associated challenges. First, the disease risk of clinical variants at the variant level is uncertain. We will quantify disease risk of clinical variants for human diseases by quantifying population-based penetrance in the exome sequences of 510,000 individuals with linked electronic health record data. This research area can refine variant interpretation. Second, little is known about the full spectrum of causal risk factors contributing to complex diseases. We will dissect the phenotypic heterogeneity of complex diseases using a novel Mendelian randomization framework. This research area can provide new insights into the heterogenous causes of complex diseases. Third, little is known about the contribution of rare coding variants on deleterious load, and its effect on human phenotypes. We will examine the interplay between fitness via the load, its constituents and human phenotypes in a very large exome sequencing dataset (e.g. 510,000 individuals) that enables capture of rare coding variants. This research area will provide insights into the bidirectional relationship between deleterious load and human phenotypes, which can inform about the genetic architecture of human phenotypes. Fourth, studies have shown that the side effects of drugs targeting genes are enriched for certain human genomic features; however, these studies have not yet translated to useful prediction of drug side effects. We will build a human genomics-guided priority index for drug side effect prediction using a drug side effect dataset and a wide array of genetic features. This research area can potentially improve selection of drug therapeutic candidates in drug development. Taken together, the research program will deliver new biological insights and useful tools that can inform prevention, diagnosis and treatment of human diseases.