Summary
Recent large-scale genetic studies have transformed our understanding of the molecules likely to be involved in causing psychiatric symptoms. This opens the door to new treatments, and personalised treatment pathways.
However, whether changes in these molecules can be detected in blood samples is underexplored; molecular markers can reflect the consequences of having a condition, rather than causing it; and markers may not be appropriate for the diversity of patients.
The datasets necessary for these analyses are now available or emerging, but they are stored in different locations and cannot easily be jointly analysed because of privacy considerations.
Project aims
In undertaking this project, you would use novel data federation techniques to discover robust and reproducible biomarkers that could serve as treatment targets or diagnostic tests.
You would test for relationships between polygenic risk scores (PRS), which reflect an individual’s overall genetic risk for a psychiatric disorder, and their blood biomarkers (proteins and metabolites) or clinical outcomes.
You would also test for correlations between pathway-specific PRS (which reflect genetic risk within particular biological pathways) and molecular/health outcomes, improving our understanding of biological subtypes. You would work with Bitfount to incorporate these analyses into data federation pipelines, to allow model fitting across multiple datasets, increasing the power and generalizability of your findings.
Working within the UKRI ImmunoMind Hub (https://immunomind.org/collaborators/) and PGC Functional Genomics group (https://pgc.unc.edu/for-researchers/working-groups/functional-genomics-w...), you would develop skills in statistical genetics, software development, proteomics/metabolomics, gene-environment interplay, and reproducible science.
Contact details
Mary-Ellen Lynall - mel41@cam.ac.uk
Opportunities
This project is open to applicants who want to do a:
- PhD