Skip to navigation Skip to main content Skip to footer

Approved research

Integrating machine learning methods and statistical genetics tools to understand pleiotropy and comorbidity in the UK Biobank

Principal Investigator: Dr Barbara Engelhardt
Approved Research ID: 45109
Approval date: December 6th 2019

Lay summary

The three main questions I would like to answer are 1) which complex traits are regulated by a similar underlying set of genotypes, 2) how do those regulatory genotypes affect the cell to influence traits, and 3) given a patient with chronic diseases, how can we use this genomic information to prioritize the most effective treatment. We are interested in many of the phenotypes in the data set, but particular in neuropsychiatric phenotypes and diagnoses and brain imaging data, chronic diseases, and sexual reproductive health. The methods I propose to develop, and the questions I hope to answer, will enable early diagnosis of disease through screening and imaging in a patient-specific way. This is very much in line with the UKBB's purpose of understanding disease and improving disease treatments. We will build advanced machine learning approaches to analyze these exciting and enabling biomedical data from UKBB. We specifically aim to build methods that address the joint regulation of complex traits, the causal cellular mechanisms underlying that trait regulation, dissecting imaging data as a specific trait type, and sequential decision making algorithms to determine the best interventions for patients undergoing long term treatment regimens.