Fraunhofer SCAI

On the Trail of Dementia with Big-Data

7. Juni 2018, 10:00 Uhr | Christina Deinhardt
Demenzkrankheit Alzheimer sorgt für schweren Schädigungen am Gehirn
Alzheimer's dementia has been known for about a hundred years, but there is still no cure.
© pathdoc |

Alzheimer's is still incurable. But the EU project AETIONOMY wants to get to the bottom of dementia: Researchers at Fraunhofer SCAI are trying to identify early disease mechanisms using model-based and big data approaches.

Parkinson's and Alzheimer's are among the most common forms of dementia and both are still incurable. Although drugs can alleviate the symptoms, they cannot stop the slow decay of the brain. Neurodegenerative diseases are a highly relevant but risky field of research for the pharmaceutical industry. So far, many billions of euros have been invested in research – but so far no effective drug for dementia has been found.

What Is the Molecular Cause of Alzheimer's?

The causes of Alzheimer's are still unknown, but why? »The classification system of medicine, on the basis of which it is still diagnosed today, dates back to the middle of the 19th century. New findings – for example from molecular biology – are not taken into account,« explains Prof. Dr. Hofmann-Apitius, head of the Department of Bioinformatics at the Fraunhofer Institute for Algorithms and Scientific Computing SCAI and academic coordinator of the EU project AETIONOMY.

But that is now about to change: The research project aims to classify neurodegenerative diseases on the basis of disease mechanisms and no longer on the basis of clinical symptoms. The researchers at AETIONOMY want to systematically record the molecular disease mechanisms and identify groups of patients in which these disease mechanisms are active. Specific therapy approaches will then be proposed for the different patient groups.

Identifying new disease mechanisms is a challenge, because the processes in the body that trigger a symptom differ from patient to patient. According to Hofmann-Apitius, there is probably not just one disease mechanism  that isthe same for all patients, but rather numerous triggers and mechanisms that can contribute differently to different individuals’ risk of developing the disease. For example, genetics and epigenetics, viral infections in adolescence, mental stress or brain trauma can increase the risk of later neurological disease.

Big Data Approach for Mechanism-Based Classification

Alzheimer's and Parkinson's disease develop on a molecular and cellular level. Therefore, AETIONOMY aims to establish a mechanism-based taxonomy or classification for both diseases by the end of 2018. »Reclassification is important if we want to find out which patient needs which active substance,« says Dr. Phil Scordis, a member of staff at USB BioPharma, the second consortium leader of the project.

The researchers at Fraunhofer SCAI solve the reclassification with a big data approach. »We collect all patient data and publications from hospitals, companies and libraries with the help of Big Data and link them together to form a knowledge base. This is now available,« says Hofmann-Apitius. The researchers around Hofmann-Apitius have developed software for this process: SCAIView enables the rapid localization of aggregated information in large amounts of text.Within minutes, millions of publications are read by the computer and converted into a coherent disease model. The result is a graphical model, a network of factors that are causally linked. According to Fraunhofer SCAI, the knowledge base now comprises a network of 3.000 nodes and 35.000 relationships. The data collected can be compared with measurement results from clinical studies and checked for matches.

Fraunhofer SCAI erstellt halbautomatisch biochemische Netzwerke, um Demenzerkrankungen auf den Grund zu gehen.
Semi-automatically generated biochemical networks help to display and recognize relationships.
© Fraunhofer SCAI

The researchers are currently creating a first taxonomy according to which patients can be classified. There is evidence that Alzheimer's and Parkinson's subgroups can be identified which are characterized by certain mechanisms. For example, frequent traumatic brain injuries, such as those that occur in American football players, can promote the development of Alzheimer's disease. The researchers have already identified 126 disease mechanisms for Alzheimer's disease and 76 for Parkinson's disease.

A follow-up project based on the research results of AETIONOMY is already being planned: According to the Fraunhofer SCAI, it is suspected that already approved drugs (including an asthma drug), which have a neurological side effect in their actual field of application, could be used for the prevention of Alzheimer's disease. Corresponding tests will start shortly.

Das könnte Sie auch interessieren

Verwandte Artikel

Fraunhofer-Institut für Algorithmen u. Wissenschaftliches Rechnen SCAI