Friday, 06-09-2019

Research

Predicting Parkinson’s Using Artificial Intelligence

Junior Research Group Leader to receive € 650,000 research funding from Michael J Fox Foundation

The Michael J. Fox Foundation is funding a Lübeck-led project for predicting Parkinson’s disease using artificial intelligence to the tune of 725,000 dollars (around 650,000 euros). The foundation, founded by and named after the US actor (known for such movies as “Back to the Future”), supports research into Parkinson’s disease, which Fox himself suffers from.

Parkinson’s disease is one of the most common neurodegenerative diseases of the brain with over 6 million affected patients worldwide, some 300,000–400,000 of whom are in Germany. Similar to other adult-onset “multifactorial” or “genetically complex” disorders, Parkinson’s is caused by an interplay of genetic and environmental/lifestyle factors.

Unfortunately, it is still not possible to predict the onset of the disease reliably in healthy individuals. “Accurate disease prediction is only possible in so-called monogenic forms of Parkinson’s where highly penetrant singular DNA variants can cause the disease. However, this form of Parkinson’s is so rare that it only plays a very minor role in the Public Health domain,” says Dr. Christina Lill, doctor and epidemiologist at the Lübeck Interdisciplinary Platform for Genome Analytics (University of Lübeck) and at the Institute of Human Genetics (University Medical Center Schleswig-Holstein, Campus Lübeck), as well as coordinator of the project.

Project partners in London, Utrecht, Heidelberg, Freiburg, Naples and the USA

“Our goal is to increase the precision of disease prediction for the more common, multifactorial form of Parkinson’s to be able to identify persons at high risk who can be included in clinical trials for early treatment or prevention of the disease,” says the researcher.

In order to be able to develop the improved disease-prediction algorithms, Dr. Lill’s group teamed up with six other European research teams and one US company to submit a research proposal to the MJFF, which has now been approved. Project partners are the Imperial College London (UK), University of Utrecht (Netherlands), Danish Society for Cancer Research (Denmark), German Cancer Research Center (Germany), University of Freiburg (Germany), University of Naples (Italy), and the company Somalogic (USA).

“The project shows great promise, not least because it utilises both clinical data and blood samples from the EPIC study, which was originally initiated to promote research on cancer and which will now also be used to study Parkinson’s disease,” explains Prof. Elio Riboli at Imperial College London, who initiated the EPIC study and is Co-PI of the new MJFF-funded project.

Data from one of the world’s largest studies on health research

EPIC (“European Prospective Investigation into Cancer and Nutrition”) began in the 1990s by collecting detailed health data from over half a million participants, including a broad array of environmental and lifestyle variables, and blood samples which are stored in a dedicated biobank. EPIC thus represents one of the largest epidemiological studies ever conducted worldwide. The health status of study participants has been and continues to be followed on a regular basis. The available data makes it possible to determine who among the study’s participants eventually developed Parkinson’s disease and which combination of factors distinguishes them from those who did not.

“In addition to the already available data, a novel high-throughput procedure developed by our project partners in the U.S. will allow us to determine the concentrations of more than 5,000 proteins in one experiment. These ‘proteome’ data will then be combined with data from other domains to develop better disease prediction algorithms,” says Prof. Lars Bertram, head of the Lübeck Interdisciplinary Platform for Genome Analytics at the University of Lübeck, from where this project will be coordinated. To achieve this aim, the international consortium will apply and develop a range of artificial intelligence techniques.

Thanks to its design, the EPIC study is ideally suited to the Lübeck project. The collaboration with EPIC was initiated after Dr. Lill was appointed as a (part time) lecturer at the Imperial College London’s School of Public Health in 2018. “Through this appointment at Imperial a plethora of new collaboration projects is made possible,” says Dr Lill. “We hope that utilising the extensive EPIC database will not only allow us to excel research into Parkinson’s but hopefully also into other common neurodegenerative diseases.”

Dr. Christina Lill won the 2015 Renate Maaß Research Award from the University of Lübeck.