Artificial Intelligence-Guided Drug Discovery for Alzheimer's Disease

Our goal is to develop an AI-guided drug discovery platform that rapidly identifies and validates new therapeutic candidates targeting tau pathology in Alzheimer’s disease.

Current AI drug design tools were built for conventional proteins and fail to account for the unique repeating fibril structure of Alzheimer’s tau. Our platform is the first to combine fibril-specific AI prediction and generative design with validation in neurons seeded with actual Alzheimer’s patient-derived tau aggregates. The experimental results will be fed back to the AI models for further fine-tuning. This integration creates a disease-relevant closed loop between computation and experiment that no existing approach provides.

This work could accelerate the discovery of drugs that slow or halt Alzheimer’s disease by identifying promising compounds far more efficiently than traditional screening. Beyond Alzheimer’s, the AI platform can be adapted to other diseases caused by protein aggregation, such as Parkinson’s disease. For the millions of patients and families affected by these devastating diseases, this research offers a new path toward treatments that do not yet exist.