Bristol
Golam Khandaker & Tom Gaunt
Academic leads
Samina Begum
Co-production lead
Ruta Margelyte & Christina Dardani
Key Researchers
Inflammatory protein targets
Objectives
1
Primary: We will prioritise inflammatory proteins that causally influence the risk of serious mental illness (SMI) and could be new treatment targets for SMI by applying Mendelian randomisation and related genetic causal inference methods to largescale multi-omic data (e.g., proteomic, genomic, and transcriptomic).
2
Exploratory: We will identify biomarkers and causal drivers for illness state, illness persistence, specific symptoms/symptom dimensions, and early-life phenotypic footprint of causal genes.
What we are doing
We have pioneered the use of a suite of genetic epidemiological methods to screen for causal effects of ~500 immune proteins on 7 psychiatric, neurodevelopmental and neurodegenerative conditions. We have updated and extended this analytic pipeline to screen a much larger number of proteins for causal effects on trans-diagnostically and sub-diagnostically defined SMI, as well as conventionally diagnosed major psychiatric conditions. Our work will include use of largescale plasma proteomic data from the UK Biobank (~3000 proteins) and other populations to identify potentially causal proteins and novel drug targets for psychiatric conditions such as depression, bipolar disorder, and psychosis.
What we are planning to do
We plan to extend our work to identify causal genetic variants, genes, and respective proteins shared across psychiatric conditions and evaluate whether these proteins might be suitable intervention targets for SMI. We will use data from large, prospective population cohorts and biobanks to understand the longitudinal and developmental mechanisms underlying the causal effects of blood proteins on SMI outcomes, and to test the effects of these genomically prioritised targets and their polygenic risk scores on clinical and psychological outcomes.
Find out more
Our approach combines Mendelian randomisation and genetic colocalisation methods to predict the causal effect of drug targets on SMI outcomes. Our approach is illustrated in this paper.
More information about Mendelian randomisation in general is provided in this 'Methods Primer'.
Our genetic prediction of drug targets is based on protein quantitative trait loci generated using the UK Biobank plasma proteomic dataset, and published in papers here and here.
Get involved
There are no opportunities at the moment. Check again soon!
