AI-driven genetic diagnosis for rare-diseases

Project Description

Exome (WES) and whole genome sequencing (WGS) are currently implemented for genetic testing of more than 7000 patients with genetic diseases per year. In order to obtain a genetic diagnosis these data are intepreted independently by a trained technician and a laboratory specialist using a wide-variety of information of the patient phenotype, known disease genes, and the genetic profile of the patient. This is a cumbersome process that can take up to weeks. The Radboudumc Department of Genetics has exome sequencing data of more than 30,000 individuals, stored as more than 500Tb of data.

We aim to develop an AI-based learning algorithm, that integrates all of this information and is able to automatically diagnose patients with genetic diseases. This will reduce (1) the amount of time spent on interpretation, and thereby also (2) reduce turn-around times for these tests, and (3) will improve diagnoses by reducing the risk of human error.

People

Christian Gilissen

Christian Gilissen

Associate professor

Human Genetics, Radboudumc

Marcel van Gerven

Marcel van Gerven

Professor

Artificial Intelligence, Radboud University

Max Hinne

Max Hinne

Assistant professor

Artificial Intelligence, Radboud University

Helger Yntema

Helger Yntema

Head of Genome Diagnostics

Human Genetics, Radboudumc