Approved Research

Discovery and validation for genome-phenome network analysis in precision medicine applications

Lay summary

Medical breakthroughs are rarely punctuated single events.  Rather they are the result of incremental advances from across different (and often seemingly unrelated) scientific disciplines.  A series of advances in deoxyribonucleic acid (DNA) sequencing and analytical methods led to the development in the 1980's and 1990's of genetic tests-and thus new clinically actionable decision points-for several diseases including sickle cell anemia and cystic fibrosis, the most common monogenic traits in Western populations.  More recently, the idea of "precision medicine" has offered the promising possibility of taking the next large leap in public health.  Some success has already been achieved on this front in fields like oncology, where survival rates have more than doubled for some cancer diagnosis types.  Nonetheless, there remain many diseases where the ability to predict onset or reverse symptoms has eluded researchers despite decades of investigative efforts.  This suggests a need for novel research methods and analytical paradigms.  The most recent wave of technological advances, especially in DNA sequencing and bioinformatics, has allowed rapid expansion of access to "multiomics" data over the last decade that allow high resolution analysis multiple aspects of an individual's molecular physiology in real time.  The challenge now lies in applying new analytical models and computational approaches to transform these data into actionable insight that will power the next generation of biomedical breakthroughs.  We have leveraged our computational resources and a multidisciplinary approach to problem solving to develop novel analytic pipelines for discovering connections between specific combinations of DNA sequence changes in a person's genome and that person's risk for developing different disease syndromes.  This precision medicine approach to gene discovery and our large and diverse patient partner population help make us uniquely positioned to extract novel and meaningful discoveries from large datasets.  The analysis we propose in this project will allow us to test whether the discoveries we have made in our own datasets can be reproduced in the UK Biobank sample.  Reciprocally, it will also allow us to make discoveries in the UK sample, which can then be reevaluated in our own independent data in-house.