The ATLAS Consortium
Dementia Frontiers fund
Decoding the Biological Terrain of Alzheimer’s Progression: Integrating Co-Pathology, Omics, and Systemic Drivers.
Go to section:
Led by
Professor Duygu Tosun (University of California San Francisco) and Dr Oliver Robinson (Imperial College London)
Research Question
Can we better predict and stage disease progression in dementia?
Locations
USA, UK and Spain
Research institutions
6
Together with
Assistant Professor Aaron Scheffler (University of California San Francisco), Professor Carlos Cruchaga (Washington University in St. Louis), Dr Marc Suárez-Calvet (Barcelonaβeta Brain Research Center), Associate Professor Gyungah R. Jun (Boston University) and Dr Lycia Neumann (Alzheimer’s Association)
Project summary
This project aims to move beyond static Alzheimer’s disease (AD) staging by developing a prognostic framework that predicts the timing of disease progression in cognitively unimpaired, amyloid-positive individuals.
While current biomarker-based approaches can identify individuals within the biological continuum of AD, they provide limited information about when clinical symptoms are likely to emerge.
The programme is centred on the hypothesis that the interval between amyloid positivity and the onset of mesial temporal tau pathology, termed amyloid-to-tau latency, represents a critical and highly heterogeneous biological intermediate that governs progression towards cognitive decline.
To test this hypothesis, the team will develop the Amyloid-to-Tau Latency Assessment Score (ATLAS), an integrated modelling framework that combines longitudinal biomarker data, neuroimaging, blood-based molecular profiling and clinical information.
The researchers will use Bayesian biomarker modelling to estimate individualised amyloid-to-tau latency trajectories, while machine learning approaches will identify co-occurring neuropathologies, including Lewy body disease, TDP-43 pathology and cerebrovascular disease, from routinely acquired MRI data.
The project will also incorporate proteomic, exosomal and circulating RNA signatures to characterise biological ageing, inflammation and metabolic dysregulation.
These multidimensional data will be integrated using advanced multi-task machine learning approaches to generate personalised predictions of progression from asymptomatic amyloid pathology to clinical symptom onset, while identifying the biological mechanisms that accelerate or slow disease progression.
Plain English summary
People with early signs of Alzheimer’s disease can develop symptoms at very different speeds. This project will investigate why some individuals remain symptom-free for many years while others experience more rapid decline.
By identifying the biological factors that influence disease progression, the researchers aim to develop a tool that can help estimate when symptoms are likely to appear and support more personalised care and treatment decisions.
Potential future impact
This research could improve our ability to predict when people with early Alzheimer’s-related brain changes are most likely to develop memory and thinking problems.
By identifying the biological factors that accelerate or slow progression from amyloid build-up to tau pathology and cognitive decline, the project may reveal new opportunities to delay or prevent symptoms.
The ATLAS prediction framework could help clinicians identify individuals who are most likely to benefit from targeted interventions at specific stages of disease, supporting more personalised approaches to care.
In the longer term, the findings could improve the design of clinical trials by enabling more precise patient selection, helping to accelerate the development of targeted therapies and preventative strategies.