Genome wide association studies (GWAS) have successfully identified thousands of loci likely affecting human health. To translate these findings into therapeutic targets and disease treatments, we need to understand the cellular context and underlying biological mechanisms through which each disease associated variant disrupts function. Large scale, information rich datasets are being generated across multiple modalities including transcriptomics from single cell RNA-seq studies, traits and phenotypes from the UK Biobank and germline genetic variation from exome sequencing studies. Here, we propose to develop methods to integrate these amazing resources towards understanding the identifying biological and cellular mechanisms that are leading to disease. The objectives will be accomplished with the following specific aims: 1) Integrate population scale biological datasets including UK Biobank and single cell transcriptomics data to construct gene modules with the goal to recapitulate biological pathways. 2) Develop a statistical framework to measure mutational burden across each of the cell type specific gene modules. Together, this research proposal will increase the power in interpreting human genetic variation and help better understand the mechanism through which they act. These methods are being developed around an IBD dataset and will derive substantial molecular information about the mechanisms driving IBD. The lessons and methodological advances from this work will be directly applicable in many complex disease contexts.
Genome wide association studies are highly successful identifying thousands of disease-associated loci but fall short at 1) pinpointing the biological mechanisms through which the variants effect disease and 2) capturing the effects of rare variants which may be having larger effect on disease outcomes. This research proposal details plans to integrate high-throughput single cell transcriptomics, population scale biobank datasets and exome sequencing from large disease cohorts to learn a prior on which genes are likely working together in which cell types and consequently improve the statistical power in interpreting rare human genetic variation. Our work will create foundational methods that can be applied broadly across many disease contexts.
