This text is part of the special notebook Innovating for better care
Alongside the tens of thousands of known genes in the human genome, there are approximately 17,000 “unidentified genomic objects” that are called “pseudogenes”. A new research chair at the University of Sherbrooke aims to characterize their role in certain pediatric pathologies.
“Since the beginning of genomics, we have called these genes that because we have difficulty seeing them and we do not know what they do. They are more or less ghosts,” explains Marie Brunet, professor-researcher at the Faculty of Medicine and Health Sciences at the University of Sherbrooke. “That’s what we want to understand,” adds the person who holds the CRMUS Chair in artificial intelligence and multiomics applied to pediatric pathologies.
This five-year chair will include artificial intelligence and clinical research at the bedside. And this, in order to unravel the mysteries of pseudogenes and find a treatment for diseases in children, such as certain forms of cancer and developmental disorders.
His team is one of the rare groups to launch into this almost unexplored field, which obviously implies a strong potential for discoveries. The objective is to be able to advance precision medicine, and more particularly certain therapies.
“Currently, medicine is very good when it comes to populations, but it is still difficult to explain why each individual reacts differently to the same therapy. However, it is impossible to achieve precision medicine if we do not know all aspects of the genome,” she emphasizes.
First, the pseudogenes
Although she plunges into the unknown, Marie Brunet nevertheless has a good idea of the steps to take to achieve the desired result.
The first will be to demonstrate the biological importance of pseudogenes. “Since 1977, they have been defined as non-functional and this label must be changed. We already have several examples which show that they can act, but they have not been characterized. »
Pseudogenes, she explains, are copies of genes known to be inactive because they are not expressed by the production of a protein, unlike genes – “we believe”, specifies the professor-researcher. However, for almost half of genetic diseases, we still do not know which genes are involved. Marie Brunet thus has reason to think that pseudogenes are involved, and that their role would also be more important than we imagine.
The second step will consist of setting up analytical processes to detect pseudogenes, identify them and distinguish them with certainty from genes, which no one knows how to do well, she adds.
This is where artificial intelligence will come into play. The human brain, explains the researcher, is simply incapable of capturing all the multidimensional relationships between parts of the genome. “To eliminate ambiguity, we need to develop algorithmic methods that will tell us “yes”, “no” or “maybe”, and which will allow us to dig even deeper into what is vague. »
AI will also intervene in the opposite direction, that is to say to analyze the “multiomic” signature of certain diseases. This term, “multiomics”, refers to the point of intersection between “genomics” (the study of the genome itself), “proteomics” (the study of proteins produced by genes) and “transcriptomics” (the how the genome is transcribed into protein). “If we can identify the deep omic signature of diseases and pseudogenes, we may be able to predict cancer recurrence,” says the researcher.
From lab to bedside
And the chair’s team of researchers will still have to use another layer of algorithms to establish the link between the omic profile of pseudogenes and that of certain cases of cancer or developmental diseases. “This is the third step of our work: demonstrating which diseases are caused by pseudogenes and how we can predict them. »
According to Marie Brunet, the clinical aspect of the work will not occur at the end of the process, but almost in parallel, because it will involve involving patients so that they participate in the research protocols.
“Even if we don’t necessarily find a therapy, we want to be able to explain certain illnesses. The simple fact of knowing does immense good to patients and those around them. There is not a single parent who does not feel guilty about their sick child,” emphasizes Marie Brunet.
There will therefore not be too many eight at the Chair to take up the challenge, specifies the researcher, but its holder is hopeful of being able to arrive at a first model. “We have already started the first two stages. We are making progress,” she believes.
The process, which involves years of trial and error, will often be frustrating, she warns. “If it were easy, it wouldn’t be research. When I have someone who gets discouraged, I send them for a trip to the hospital, to remember why we do this, to improve the outcome of patients. »
This content was produced by the Special Publications team at Dutyrelating to marketing. The writing of the Duty did not take part.