Discovering Biology for Neuropsychiatric Diseases Through Omics Studies on Comorbidities

点击翻译按钮获取中文摘要

Abstract: We propose in this application to use truly unique resources available to the Vanderbilt University research community to identify and characterize genetic risk factors for neuropsychiatric disorders. Our overarching hypothesis is that co-morbid phenotypes that cut across neuropsychiatric disorders can be used to identify more homogeneous genetic risk factors that will also be cross-cutting for neuropsychiatric diseases. To address this hypothesis, we will harness the long-standing strengths in neuroscience at Vanderbilt including extensive expertise in conducting in vivo and in vitro experimental validation studies, the strong team of investigators with long-standing research programs in key co-morbid phenotypes and neuropsychiatric disease, and our track record in developing and applying novel integrative approaches for genome investigation. The clinical data warehouse at Vanderbilt is called the Synthetic Derivative (SD), and contains continuously updated electronic health records (EHR) on more than 2,500,000 individuals. DNA samples are available on more than 217,000 of the individuals in the SD through BioVU, the biobank at Vanderbilt University. Individuals with more longitudinal data some going back as long as 20-30 years have been prioritized for genome investigation, and genome interrogation (GWAS or whole genome sequencing) will be available on > 120,000 of these subjects in 2018. The SD provides unprecedented power for characterizing cross-cutting comorbidities for neuropsychiatric disorders, and the large number of BioVU samples with genome interrogation coupled with the novel analytic approaches we have devised to optimize genome investigations in BioVU create a dynamic engine for discovery research. Our specific aims are to: 1) Use EHR data on more than 2,500,000 individuals to investigate the relationship between neuropsychiatric disorders and comorbid phenotypes shared among multiple of these disorders; 2) Use the novel PrediXcan approach to identify genes for which genetically predicted expression is significantly associated with neuropsychiatric disease, neuropsychiatric disease plus comorbidity, or comorbidity for more than 120,000 samples in BioVU; and 3) Prioritize genes for validation using improved network and pathway analyses, and then experimentally validate genes implicated in neuropsychiatric and comorbid phenotypes.