Project 8: Predicting patient trajectories and outcomes from national level data

Project description

A polygenic risk score is a single value estimate of an individual’s genetic propensity towards a trait or disease. In the last decade, numerous methods have been published that aim to better polygenic risk score calculation. These methods may work on some specific dataset or trait, but whether they are generalise well is debatable. To incorporate polygenic risk scores in clinical practice, they need to be robust, and therefore there is a need to enhance the score’s sensitivity and specificity. One way to do this is to incorporate clinical observations from electronic health records and genetic info and apply machine learning algorithms to assess risk for a person to develop a particular disease. Therefore, we need to find conditions that occur frequently enough in Estonian electronic health records data to use them in models or as proxies for assessing disease progression. We can then build a baseline predictor that we use later to validate that our enhanced risk score predictor with clinical inputs works significantly better.

Secondments

1. Polygenic risk score prediction for Depression
   Host: Max Planck Institute of Psychiatry
   Planned date: February 2021 – April 2021 Postponed due to Corona
2. Improving polygenic risk score through high-dimensional patient trajectories
   Host: UC3M
   Planned date: March 2022 – May 2022