Skip to navigation Skip to main content Skip to footer

Approved Research

Machine learning approaches for polygenic disease modelling

Principal Investigator: Professor Samuel Kaski
Approved Research ID: 77565
Approval date: November 9th 2021

Lay summary

Modern machine learning methods, such as neural networks, have proven highly successful at representing complex relationships in large datasets. When applied in the health domain, machine learning methods can integrate data from multiple sources to facilitate a better understanding of the interactions between genetics, lifestyle/environmental factors and disease. This research will focus on the development of new machine learning methods for modelling polygenic diseases, which are diseases caused by the combined effects of multiple genes.

The algorithms developed in this research can aid prediction of disease risk and personalised treatment. We aim to achieve this through the development of machine learning methods integrating genetics data with additional information from longitudinal health records and blood biomarkers, which will help characterise the heterogeneous nature of disease. Our research will focus on highly polygenic diseases, including cardiovascular diseases, breast cancer, prostate cancer and type 2 diabetes.

This research will take place over several years, with part of this work occurring in conjunction with the INTERVENE research consortium, which aims to integrate artificial intelligence and human genetics to develop tools for disease prevention, diagnosis and personalised treatment: