Alzheimer’s disease is a disorder characterized by abnormalities in several different biological areas.
A study carried out by HBHL-funded researcher Yasser Iturria-Medina and his research group, in the Department of Neurology and Neurosurgery at The Neuro, represents the first and largest multimodal data collection effort to date for Alzheimer’s disease. It includes two independent, large-scale, post-mortem and in vivo cohorts covering the entire disease spectrum.
By developing a new machine-learning method, the group was able to assemble different layers of molecular data into the first personalized multi-level index of Alzheimer’s dementia progression. It can predict the severity of the abnormalities affecting the nervous system and identify both disease-progression stages and distinct disease subtypes at the molecular level.
Iturria-Medina’s research group identified three molecular-based subtypes of Alzheimer's that thoroughly describe the differences in physiological and clinical presentation. These subtypes clearly show patterns of changes in DNA methylation, RNA, proteins and metabolites, neuropathologies, cognitive changes and brain-cell types. These patterns are identifiable in both the brain and blood of the patient.
Alzheimer's dementia is a complex disorder, and the group's research shows that the genetic variations that predispose patients to each disease subtype can predict distinct patterns of cell-type changes in the brain. This suggests that each known Alzheimer’s variant has a unique influence on the mechanisms of the disease.
The initiative was the result of a multi-center collaboration, including the Neuro (Montreal Neurological Institute-Hospital), McGill University, Rush University Medical Center, the Douglas Research Centre, Pacific Northwest National Laboratory, Columbia University Irving Medical Center and Yale University. This work represents the most significant effort so far to classify Alzheimer's disease from a multi-level molecular perspective.
The study’s observations will pave the way to an individually-tailored, multi-system molecular classification of Alzheimer’s disease that may help identify clear targets for preventive treatments or medications.
All the analytical tools developed for the project are shared with the community through user-friendly software and code available on the lab website.
Read the complete paper on Science Advances.
