Skip to navigation Skip to main content Skip to footer

Combining Genetic and Environmental Information to Predict Disease Risk

Combining Genetic and Environmental Information to Predict Disease Risk

Principal Investigator: Dr Bjarni Vilhjalmsson
Approved Research ID: 58024
Approval date: March 2nd 2020

Lay summary

The aim of this project is to develop a new computational approach to predict disease risk of an individual based on various information, including the genome of the individual, family history, medical history, and other information. Such predictions are useful in several different scenarios, both in clinical and research applications. In clinical settings risk predictions are routinely used to identify at-risk individuals, for example, BMI and blood measurements are currently used to assess risk for developing heart diseases. Similarly, genetic tests are commonly used to estimate risk for developing breast cancer. In this context, we believe our work can help improve the accuracy of these risk assessments. However, for us the main motivation for improving risk prediction stems from research applications, and not clinical application. In research settings, understanding risk for developing a specific disease can provide insights into what actually causes the disease, and hopefully, how we can prevent it. Similarly, the predicted disease risk can be used to study the relationship with other diseases and their underlying biology. 


The UK biobank dataset is uniquely suited for carrying this work out, as it is an exceptionally rich data resource. It will allow us to consider a wide range of parameters when optimising our predictor, including the genome, health records, and other information when predicting disease risk for an individual. Moreover, using the genome of an individual we can identify other relatives in the UK biobank data, and consider their information as well when predicting the disease risk.  The family information can tell us something about the environment that the individual in question lives in.


In terms of specific applications, we will apply the method to study mental health. By predicting risk for developing psychiatric disorders and related medical conditions, we hope our research will eventually provide valuable insights into the underlying causes of these debilitating conditions.