Combining genetic and functional genomic data to reveal underlying mechanisms of human disease

The unprecedented access to genome-wide association (GWAS) signals for a wide variety of human diseases opens up new opportunities to improve understanding of the underlying mechanisms of disease and identify new therapies.

Previous work in the Baillie lab has demonstrated that many biological pathways can be detected from expression patterns in high-resolution transcriptomic data. They interrogated shared activity patterns arising from regulatory regions containing variants associated with inflammatory bowel disease, where re-analysis of two large GWAS studies revealed two distinct groups of variants associated with both Crohn's disease and ulcerative colitis. This discovery may indicate two distinct mechanisms underlying each disease. Alternatively, it may indicate the existence of two distinct endotypes of each condition. It is at least plausible, that patients with a preponderance of 'immune' or 'epithelial' genetic variants will respond differently to immunomodulatory therapies.
The project will develop the computational and statistical tools to detect and validate mechanistic relationships between diseases for which GWAS data are available (340 diseases at the time of writing).
This will form several distinct stages which will overlap in time during the course of the project:
1. Detection of mechanistic pathways. Optimisation of coexpression methodology for high-performance computing and incorporation of data from different sources including GTEx, Roadmap Epigenetics and ENCODE.
2. Development of methodology for evaluation of disease-disease interactions, including linkage disequilibrium score regression, genomic correlation and coexpression analysis.
3. Application to existing and ongoing GWAS studies. Re-analyses of published and ongoing GWAS studies, including UK biobank, will be performed to detect distinct biological pathways underlying clinical phenotypes. Candidates will be chosen for further validation, in large population studies or clinical trials where genotyping data are available. Biological validation of specific mechanistic hypotheses will be performed in genome-editing experiments collaboration with wet-lab scientists in the Baillie lab (myeloid cells, endothelial cells) and others (hepatocytes, epithelial cells, iPSC-derived primary cells).