Diagnosing patients with rare diseases through a combination of artificial intelligence, smart computing power and genetics. To make this innovative project by Radboudumc, BioProdict and Vartion possible, the European Regional Development Fund (EFRO) and the province of Gelderland are supporting the project ‘Diagnostics-in-3D’ with more than half a million euros.
Many little ones make one big one. This also applies to rare hereditary diseases. One rare disease, as the word suggests, is not very common. But add up the roughly 7,000 known rare diseases and it soon amounts to about eight percent of the population. In recent decades, global efforts traced the causes of all these rare diseases. At the moment, this is possible for about half of the people with such a rare disease. So not for the other half. These patients do not know what exactly causes their illness.
Radboudumc, together with the companies Bio-Prodict and Vartion from the Nijmegen region, aims to change this with the project "Diagnostics-in-3D". This project is financially supported by the European Regional Development Fund (EFRO), more specifically by OP Oost, thereby contributing to a structural strengthening of the economy in the provinces of Gelderland and Overijssel. Martijn Huijnen, theme Metabolic Diseases, in the Radboudumc will lead the project. The starting point is the genetic code of humans. That code consists of approximately three billion letters. Many letters are the same, but a significant portion differs from person to person. Put one genome (the total of all genetic information of a human) next to the other and you will see many differences.
The importance of a good diagnosis
Regional cooperation - also aimed at strengthening the innovative, economic structure - is obvious. Huijnen: “The Nijmegen-based company Bio-Prodict has developed software that has automated the prediction of the effect of mutations on proteins. By using anonymous patient data and AI from Radboudumc, major steps can be taken in predicting the effect of mutations on the 3D structure of a protein, and thus on its function. For this, the representation of the 3D protein data, the deep learning architecture to be used and the computing power will be optimized. Malden-based Vartion is a specialist in the latter area. This collaboration of a knowledge institute and companies should ncrease the number of diagnoses for rare diseases. This is important, not only because a correct diagnosis is essential for the development of a treatment, but also because it may affect decisions about future pregnancies. ”
Which difference has an effect?
Martijn Huijnen: “The problem in the search for the cause of a rare hereditary disease is that it is not easy to indicate which difference in the genome exactly causes the disease. Suppose you find 1000 genetic differences in someone compared to the standard, which of those differences exactly makes you sick? Most differences have no effect whatsoever. As proteins do the work in our body, it is interesting to see whether a genetic difference also leads to important changes in the associated protein, making it less effective. That could be the cause of the rare disease. ”
A protein is a string of amino acids that has been carefully folded. That folding - organic origami you could say - is essential for the function of the protein. It provides the "hands and feet" with which a protein makes contact with other proteins and can do its job. Any genetic difference that causes a change in the shape of the protein could well be the cause of the rare disease. “In this way, the cause of rare diseases is increasingly - and successfully - searched for,” says Huijnen. "In our "Diagnostics-in-3D "project, Bio-Prodict, Vartion and Radboudumc now want to automate and accelerate this approach. In doing so, we mainly focus on the rare diseases that cause intellectual disabilities, a subject in which we have already built up a great deal of expertise in Nijmegen.”