Genome4Brussels: AI, Genomics & Rare Diseases


GEMVAP, 2021, G4BXL / Monday, May 31st, 2021

In 2019, the 101 Genomes Foundation, theInteruniversity Institute Of Bioinformatics Brussels (IB2), theULB Center of Human Genetics (CHG) and theULB Machine Learning Group (MLG) participated in a call for projects launched by Innoviris with the "Genome4Brussels" project. Within the framework of this project, they decided to create together an ecosystem that will optimise the development of bioinformatics tools for genome analysis and facilitate the transfer of the innovation and knowledge acquired during the project to the public.

Project

Genome4Brussels. Genome4Brussels is a joint project that aims to create an ecosystem in the Brussels region that combines :

  • optimal conditions for hosting and sharing genomic data led by patient representatives (Fondation 101 Génomes);
  • medical and genomic expertise (CHG-IB2) and ;
  • expertise in bioinformatics and artificial intelligence/IA (IB2-MLG).

This ecosystem will allow the emergence of a research platform dedicated to the development of bioinformatics tools developed through "transparent" AI (White Box) to assist physicians and researchers in the field of rare diseases.

In the framework of the Innoviris call: a research platform is set up, two bioinformatics tools are developed and the conditions for transferring these technologies to the public are created.

Partners

101 Genomes Foundation. In 2017, the 101 Genomes Foundation embarked on a quest to find genomic superheroes whose genes protect them from the effects of certain rare diseases. To carry out this quest, Fondation 101 Génomes is starting by developing a database from which scientists can explore the human genome for protective (or aggravating) genes that explain the variability of rare diseases. Such a discovery would provide better diagnoses and allow for new treatments that replicate the protective (or limit the aggravating) effects identified. The 101 Genome Foundation's pilot project is dedicated to Marfan syndrome. This pilot project is supported by several European patient organisations and is led by leading scientists.

IB2-CHG. L’Interuniversity Institute Of Bioinformatics Brussels (IB2) and theULB Center of Human Genetics (CHG) are jointly dedicated to the study of rare diseases. Their common mission is to improve the diagnostic quality of genetic tests, in order to improve clinical follow-up, treatment and therapeutic advice. Their research areas are part of this overall objective, and consist of identifying the genetic origin of rare diseases that are still insufficiently understood. IB2 and CHG are developing bioinformatics tools to explore and analyse patients' genomic data in an innovative way and thereby improve the quality of genetic/genomic testing.

IB2-MLG. L’ULB Machine Learning Group (MLG) specialises in computer science, artificial intelligence, bioinformatics, computational biology, genetics, molecular biology and medicine. In the medical context, IB2 and MLG use methods associated with statistical scores that create a transparent "White-Box" model providing explanations of the decision made by the bioinformatics tools they develop with AI. It is indeed unthinkable to be satisfied with a basic AI approach (which provides results without a "Black-Box" explanation) in the context of the development of bioinformatics tools intended to assist clinicians.

Fair Genomics (FairGX). In the course of the procedure, Innoviris asked the consortium to set up the Fair Genomics to accompany the project and, in so doing, enable a transfer to citizens of the technologies and innovation developed through Genome4Brussels. Fair Genomics is wholly owned and controlled by Fondation 101 Génomes . 

Ecosystem. The partners intend to set up a virtuous circle to fuel and fund research and advance science.

Marfan syndrome research funded from 2020 to 2023

Innoviris. The project 'Genome4BrusselsThe 'Marfan Syndrome Pilot Project' has been accepted by Innoviris at the end of 2020. Innoviris will fund part of the project which will allow the IB2, CHG and MLG teams to work for three years (2020-2023) on the Marfan syndrome pilot project.

 


 

In May 2021, Innoviris published an article about Genome4Brussels which can be consulted by following this link: https://innoviris.brussels/fr/stories/ludivine-verboogen-et-romain-alderweireldt
The article is also reproduced here:

Ludivine Verboogen and Romain Alderweireldt

« There is a paradox in rare disease research. Funding for rare diseases is still problematic. But at the same time, research on rare diseases has led to significant advances in other more common diseases. The added value of this research goes far beyond rare diseases. If we analyse things very coldly, we can say that we are missing out on enormous economies of scale by funding little or poorly this research on rare diseases« . 

Romain Alderweireldt and Ludivine Verboogen only need a few words to make clear the meaning and significance for society of their joint commitment to Fondation 101 Génomes. Their ambition? To provide scientists with a bioinformatics platform to advance in the field of genomic research, one of the keys to better understanding rare diseases.

The vagaries of life create a wonderful impetus for the public good

The story of the foundation actually begins on 3 September 2015 with the birth of Aurélien, the son of Ludivine and Romain.

The joy of the first few days was soon followed by concern: the doctors suspected Marfan syndrome, a rare disease that can severely affect connective tissue, with consequences including heart disease.

On 4 August 2016, 11 months later, a genetic analysis confirmed the bad news: Aurélien is a carrier of a spontaneous mutation of the FBN1. He does have a severe form of Marfan syndrome.

A late diagnosis and tools to be invented

This is a spontaneous and rare variant of Marfan syndrome. This played a role in the 11-month delay in making the diagnosis: the reference centre requested by the geneticist who was following Aurelien to carry out a genetic analysis was still using gene-by-gene analysis techniques, whereas genomic analysis techniques existed "explains Ludivine. « And it was during the many discussions we had with this geneticist, Professor Guillaume Smits, that we realised that the clinic lacked the physical and bioinformatics tools to take advantage of the advances in genomic research, both in terms of diagnosis and the development of new treatmentssays Romain.

« Six years later, it is possible to sequence the entire human genome and identify mutations in all our genes at an increasingly affordable cost. But the initial cost of the machines is still very high and genome sequencing has not yet become routine in research or clinical practice. A sequenced genome generates such a large volume of data that it is absolutely impossible to process it manually: one genome corresponds to nearly 300 gigabytes of data! You need algorithmic tools, artificial intelligence and a highly specialised team to get anything out of it. This tandem between ICT and biological research has a name: bioinformatics. »

Turning disarray into positive energy

For this legal couple, the realisation that they needed to invest in genomics and bioinformatics was a trigger. Ludivine and Romain turned their confusion into positive energy with one goal: to advance research and the clinic, and to contribute to the emergence of treatments that will help Aurélien and other children with rare diseases.

In search of the protective gene

Romain started devouring scientific studies. That's how he came across the Resilient Project's reanalysis of nearly 600,000 exomes (the coding part of the genome). Many rare diseases, including Marfan syndrome, are caused by pathogenic mutations, or if you prefer 'abnormalities', in the genesexplains Ludivine.

« By re-analysing the data available to it, the study identified 13 adults who carried a genetic defect that should have made them very ill or even killed them as children. However, these 13 people had reached adulthood "Romain continues. « Why is this? The presence of other genes with a protective effect that counteracts the disease could be one explanation. Studying the genome of these people could therefore be the key to better understanding this protective mechanism. And this understanding could lead to the development of drugs or new treatments, for example based on the proteins produced by these genes. »

Ludivine: " Romain then decided to replicate this approach in the specific context of Marfan syndrome. And the amazing thing happened: he managed to find, in the global genomic reference database (gnomAD), carriers of pathogenic mutations that were supposed to have a severe form of Marfan syndrome and apparently did not. It became clear to us that if a protective gene could be identified, its effects could be replicated to develop a treatment. »

The Fondation 101 Génomes and its pilot project

And in 2017, the Fondation 101 Génomes was created on the initiative of Ludivine and Romain.

The objective? To develop a bioinformatics platform to host genomic data and cross-reference it with the medical records of rare disease patients and "control" volunteers to better understand the interactions between genes within the genome.

Soon a pilot project was launched: the 101 Marfan Genomes project, which was the establishment of a cohort of 101 Marfan syndrome patients as a basis for research work. It quickly became clear that 101 genomes would only be a start, and that they had a crucial need to surround themselves with knowledgeable scientists.

The scientific committee at the helm of the Foundation is made up of world leaders in genomics, Marfan syndrome and algorithms, including several Belgians. And these scientific experts admire the work done by Ludivine and Romain, which they consider to be an unprecedented example of patient participation in scientific research.

« The work focused on the creation of a cohort of patients carrying the same mutation to limit confounding factors (GEMS project), and in parallel on the creation of an algorithm designed to confirm (or refute) the pathogenic nature of mutations in the FBN1 gene. On this basis, we hope to explore the entire genome to identify modifier genes (protective or aggravating) for Marfan syndrome. The first results are promising "says Romain. « And the intervention of Innoviris helped to give this work a boost. »

GENOME4BRUSSELS, an ecosystem at the crossroads of 3 disciplines

Because the Innoviris call for projects 'From therapeutic medicine to predictive medicine: Prediction, Prevention, Identification' was received 5/5 by Romain, Ludivine and the Foundation.

Aiko Gryspeirt, Scientific Advisor at Innoviris: " Too little attention is still paid to rare diseases. As a result, there is also a lack of funding for such research, both disease-specific research and treatment. Patients often spend months and years going from one specialist to another before getting a diagnosis. With GENOME4BRUSSELS, Ludivine and Romain have launched a fantastic initiative to make it easier to understand rare diseases by collecting genetic data and sharing it with medical researchers and bioinformaticians in Brussels. This is an invaluable project that can have a significant impact on the lives of people living with a rare disease. »

"GENOME4BRUSSELS is a project financed by Innoviris in which the Foundation is involved, as well as the Interuniversity Institute of Bioinformatics in Brussels, the Center of Human Genetics and the Machine Learning Group, all three of which are at ULB. The objective is to create an ecosystem in Brussels dedicated to medical, genomic and bioinformatics expertise. This support has enabled our pilot project to move forward and combine our expertise to identify protective genes that will pave the way to better help patients".,
concludes Romain.

 

 

Leave a comment

This site uses Akismet to reduce unwanted. Learn more about how your feedback data is used.