The key appeal of polygenic risk scores (PRS) is the provision of individual-level estimates of genetic liability to complex disease. These proxies of genetic liability enable a raft of applications across clinical and basic research settings. However, while PRS are set to play a pivotal role in the future of biomedical research, their present formulation is suboptimal since it fails to directly account for substructure in genetic disease risk. The overarching goal of our proposal is to introduce a new generation of pathway-specific PRS, informed by biological function. Rather a single genome-wide PRS for each individual, they will have a set of k PRS over k pathways. Pathways will be defined according to multiscale integration of ?omics data, exploiting co-expression networks, the transcriptome and the epigenome. The key deliverable from this project will be the production of a powerful and comprehensive pathway-specific PRS computational tool, PRSet, informed by biological function. The rationale is that PRS calculated for individuals by aggregating the effects of all risk variants genome-wide, results in a loss of vital individual-level information. Providing pathway-specific estimates of genetic liability, computed in a scalable, statistically rigorous way, informed by latest multi-omic data, could enable researchers to better decompose heterogenous complex disease, identify key pathways that explain overlap or differences among disorders, and explain problems of portability of PRS between and within populations. Applying our pathway-specific PRS tool, we seek to stratify patients into more homogenous subgroups by their liability over key pathways. We will use PRSet for stratification in three ways: (i) stratifying within SCZ/BiP, testing if liability over different pathways forms multiple routes to disease, (ii) differentiating between SCZ and BiP, testing if key pathways differentiate these highly overlapping disorders, (iii) testing whether variation in treatment response can be explained by pathway liability. Such stratification could help explain past successes, failures and adverse-effects in clinical trials, and provide new therapeutic targets tailored to subsets of patients. Our proposal is significant because the burgeoning application of PRS means that any advance in the PRS approach will have immediate, high impact across psychiatric research. Pathway-specific PRS could open-up routes to hypotheses that cannot be answered by genome-wide PRS. If PRSet reveals that genetic liability is more stratified than presently modelled, then this would call for a focus on pathways and their multi-omic integration, paving a new path towards precision medicine. Our proposal is innovative because we develop the first pathway-specific, function-informed, PRS tool, we propose that disease risk may be influenced by multiple genetic liabilities, and we stratify patients according to pathway-specific genetic risk for the first time. In conclusion, our proposal delivers a tool for the field to perform powerful pathway PRS analyses, better understand genetic liability to disease, and which may offer a more direct route to precision medicine.

Public Health Relevance

Polygenic risk scores (PRS) are set to play a key role in advancing our understanding of the etiology and treatment of disease. While the application of genome-wide PRS is burgeoning across biomedical research, we propose that pathway-specific PRS have even greater potential to provide etiological insights and will help pave the way to precision medicine. Here we introduce a new generation of pathway-specific polygenic risk scores, informed by biological function, and deliver a comprehensive computational tool for their analysis, applying it to perform patient stratification in schizophrenia and bipolar disorders.

National Institute of Health (NIH)
National Institute of Mental Health (NIMH)
Research Project (R01)
Project #
Application #
Study Section
Genetics of Health and Disease Study Section (GHD)
Program Officer
Arguello, Alexander
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Icahn School of Medicine at Mount Sinai
Schools of Medicine
New York
United States
Zip Code