A research consortium co-led by the Buck Institute for Research on Aging and Georgetown University has received an award of up to $51.7 million from the Advanced Research Projects Agency for Health (ARPA-H) to predict how disease-causing proteins misfold before irreversible damage occurs. The four-year project, called team SEEDS, is meant to produce new ways to detect, slow and treat neurodegenerative diseases and cancer, with Alzheimer's disease and Ewing sarcoma as the first targets.
The proteins at the center of the work are intrinsically disordered proteins, or IDPs. Most proteins fold into fairly stable three-dimensional shapes, but IDPs can take many different forms, and conditions inside the cell such as acidity, crowding, metals and chemical modifications can shift which forms they adopt. That flexibility helps healthy cells function, but it also makes these proteins hard to study and hard to target with conventional drugs, which are usually designed to fit stable structures and binding pockets. Proteins of this kind are linked to Alzheimer's, ALS and some cancers.
Current tools can capture snapshots of these proteins, often only after they have already clumped together. The SEEDS team wants to move from describing them one at a time to predicting how they behave from their sequence and surroundings. Zhou, an assistant professor at the Buck Institute, notes that AI tools such as AlphaFold have advanced protein science but rely on large amounts of structural data that do not yet exist for disordered proteins. The team plans to pair computer predictions with laboratory measurements taken under varied conditions.
Two linked research groups will generate that data. The Buck Institute's SHIFT effort will map the context-dependent shapes of disordered regions across many human proteins, while Georgetown's INTREPID effort will examine tau and EWS::FLI1 in depth, the proteins implicated in Alzheimer's and Ewing sarcoma respectively. The data will be used to train foundation models that predict how these proteins behave. The work is part of ARPA-H's BIOGAMI program, which focuses on the earliest stages of protein dysfunction, before lasting cell damage occurs.
For patients and families, the stakes lie in whether these earliest molecular shifts can be spotted and redirected. The source material frames the project as a bet that prediction could lead to intervention, though it does not describe any treatment or diagnostic that is available or close to clinical use.