MONTREAL, November 25, 2022- A research team led by Dr. Despoina Manousaki at the CHU Sainte-Justine Research Centre is embarking on a project that will help improve genetic prediction of type 1 diabetes (T1D) in non-Europeans. The researcher, who is also an assistant professor in the Departments of Pediatrics and Biochemistry and Molecular Medicine at Université de Montréal, and a pediatric endocrinologist at CHU Sainte-Justine, has just been awarded a grant from JDRF to conduct research aiming to develop a better trans-ancestral genetic risk score for this disease.
A chronic, autoimmune, and complex disease, T1D, occurs when the pancreatic beta cells that secrete insulin—a hormone allowing our bodies to absorb glucose for energy—are mistakenly destroyed by the immune system. T1D usually occurs in children and young adults, and its causes are not fully understood. However, it is known to be a polygenic disease, which means that T1D risk is influenced by the combined action of more than one gene.
Polygenic risk scores (PRS), which estimate the risk of an individual to develop a specific polygenic disease based on genes from the entire genome, have yielded promising results in predicting development of T1D in Europeans. In combination with autoantibodies and clinical data, these PRS could help select candidates for prevention studies using new medications. However, PRS, typically developed using European data, perform poorly in diverse ancestral populations due to differences in the genetic architecture of polygenic diseases among ancestries. This is particularly important for T1D, since human leukocyte antigen (HLA) genes, which explain the largest part of T1D heritability, differ significantly among patients from different ancestries. With the project Development of a trans-ancestral genetic risk score for type 1 diabetes, the objective of Dr. Manousaki’s team is to develop a trans-ancestral PRS for T1D—one that can perform almost equally well in different ancestries.
“Disease risk prediction is key for prevention efforts for type 1 diabetes. Identifying children who are at high risk and enrolling them in prevention trials has the potential to change the life trajectories of these patients. Although accurate genetic prediction of type 1 diabetes is possible for individuals of European ancestry, the performance of genetic prediction algorithms drops significantly in non-Europeans,” explains Dr. Despoina Manousaki. “Similar to other polygenic diseases, restricting genetic risk scores for T1D to Europeans limits their use in research and clinical practice in Canada and elsewhere in the world. This could exacerbate public health inequities. Our goal is to tap into the potential of assembling genetic samples from diverse ethnic cohorts of T1D patients and controls and of using novel machine learning approaches to create PRS for T1D that can perform equally well in different ethnicities.”
Machine Learning to Build on Existing Data
To do this, researchers will apply machine-learning methods allowing the use of large available European data while incorporating information from smaller ethnic-specific genetic data from T1D patients and controls. The team will test and compare the performance of these trans-ancestral PRS on Africans, Indians, Latinos, South-East Asians and Chinese, to that obtained in Europeans. If the project is successful, it will be a major step in the global application of PRS in T1D’s risk prevention, allowing better selection of candidates for prevention trials.
– 30 –