Using genetics to stratify patients and improve prediction of clinically relevant outcomes in psychiatry and neurology (360G-Wellcome-222811_Z_21_Z)

This project investigates the genetic basis of psychiatric disorders and neurological diseases with three key goals. First, to gain insight into the molecular mechanisms underlying genetic associations with these outcomes through the integration of genome-wide association study (GWAS) summary statistics with a range of newly released functional genomic annotations. These include variant-level annotations such as the effect of genetic variation on gene expression, methylation and chromosome structure within relevant tissues and cell types, as well as gene-level annotations such as expression profiles across tissues, cell types and developmental stages. Second, functionally informed polygenic scores, such as imputed and observed gene expression risk scores, will be stratified by enriched functional genomic annotations and used for clustering patients into aetiological subtypes, which will then be characterised using sociodemographic and clinical variables. Lastly, this study will evaluate the utility of machine learning models containing functionally informed polygenic scores and cluster membership information for prediction of diagnosis, prognosis and treatment response. I will use data from a range of ancestral populations to improve the accuracy of genetic predictors in all populations. Together this project can inform novel drug development, identify clinical subgroups, and promote personalised medicine through improved prediction of clinically relevant outcomes.