Population Stratification in Risk Prediction Models

• Applicant: Dr. Damian Gola
• Subject Area: Epidemiology and Medical Biometry/Statistics
• Term: from 2019 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 428069041

An important goal of precision medicine is to estimate the individual risk of common diseases in order to enable effective prevention and early intervention. Current risk prediction models are often created using data from a specific population. Recent results, however, show that even minor differences in the underlying genetic architecture of different populations can distort individual risk predictions. This is of paramount importance if the individual comes from a different population than the model was created. This can lead to unnecessarily increased prevention efforts if the risk is considered too high, or to the lack of prevention efforts if the risk is considered too low, which in turn can lead to a lower quality of life. We are facing this situation, especially in Europe, because migration from one European country to another has never been easier than it has been these days.In this project, I first want to examine how sensitive current methods are for creating risk prediction models for different European populations. In a second step, I want to investigate how better risk prediction models can be constructed considering the underlying population structure.

• DFG Programme: Research Fellowships
• International Connection: Estonia
• Hosts: Professorin Dr. Krista Fischer, Ph.D.; Dr. Mait Metspalu, Ph.D.