What is it about?
It’s about a new way to model how Alzheimer’s disease spreads through the brain. More specifically: In Alzheimer’s, harmful proteins (like amyloid-beta) become misfolded and then spread from one brain region to another. Most past research assumes this spreading happens one connection at a time, using simple brain network models. But real biological spreading often involves group interactions (several regions influencing each other together). This study uses a more advanced type of network—called a higher-order network—that can capture these group effects. Using this newer model (a simplicial complex contagion model), the researchers were able to predict where the misfolded proteins would spread over two years, more accurately than earlier studies.
Featured Image
Photo by Robina Weermeijer on Unsplash
Why is it important?
1. Better understanding of how Alzheimer’s actually progresses Alzheimer’s spreads through the brain in a very complex way. Traditional models assume protein spread happens only one connection at a time, but biology often works through cooperative, group-level effects. Using higher-order networks means we can model the disease more realistically, which helps scientists understand why different patients decline differently. 2. Improved predictions for diagnosis and monitoring If we can better predict where misfolded proteins will accumulate next, we can: Detect Alzheimer’s earlier Track progression more accurately Identify patients at higher risk of faster decline This kind of prediction is valuable for doctors and for designing personalized follow-up. 3. Better tools for clinical trials and drug development Most Alzheimer’s drugs aim to stop or slow protein spread. If we have a model that predicts this spreading more accurately: Trials can identify who will benefit most Researchers can evaluate whether a drug is slowing spread We get clearer signals from clinical data, reducing trial failures
Perspectives
1. More accurate forecasting of disease progression Because higher-order models can capture complex spreading patterns, they could be used to create patient-specific predictions of how Alzheimer’s will evolve. This could support personalized care, risk profiling, and early interventions. 2. Integration with biomarkers and imaging These models could be combined with: PET imaging CSF biomarkers blood-based biomarkers MRI structural and functional data This would allow richer, multimodal models of protein propagation, improving predictive accuracy. 3. Application to other neurodegenerative diseases The same spreading principles apply to: Parkinson’s disease (α-synuclein) ALS (TDP-43) Frontotemporal dementia (tau variants) Higher-order networks could reveal shared or disease-specific patterns of protein spread across disorders. 4. Better mechanistic understanding The approach may uncover: cooperative or group-level mechanisms in protein aggregation non-linear “tipping points” where spreading accelerates why some regions resist spread while others are vulnerable This supports biological discovery beyond what pairwise graphs can explain. 5. Clinical trial design and drug evaluation More precise simulations can help: identify optimal trial participants predict treatment effects on protein spread evaluate whether a drug slows or alters propagation pathways This could significantly reduce trial failures, a major problem in Alzheimer’s research. 6. Development of hybrid AI + network models The method provides a foundation for combining: mechanistic models machine learning generative AI graph neural networks and topological deep learning to create new classes of disease-progression models. 7. Longer-term longitudinal studies Expanding beyond a 2-year window will clarify: how early spreading begins how propagation influences cognitive decline whether higher-order models remain superior over longer timelines 8. Toward actionable healthcare tools If validated clinically, such models could become part of: clinical decision-support systems diagnostic pipelines risk prediction tools digital twins of patient brains providing real-world impact.
Dr. Alessandro Crimi
Read the Original
This page is a summary of: Modeling the Spread of Misfolded Proteins in Alzheimer’s Disease using Higher-Order Simplicial Complex Contagion, July 2025, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/embc58623.2025.11253716.
You can read the full text:
Contributors
The following have contributed to this page
