About This Project

We combine molecular dynamics simulations, machine-learning guided molecular design, and docking to characterize amyloid-beta primary nucleation intermediates. Structural models of monomers and early oligomers enable identification of druggable sites, while virtual screening prioritizes small molecules that may disrupt initial aggregation. This computational approach aims to elucidate early pathogenic mechanisms and inform therapeutic development for Alzheimer’s disease.

Ask the Scientists

What is the context of this research?

Alzheimer's disease impacts millions of people worldwide, and one of its hallmark features is the buildup of amyloid-beta (AB) proteins in the brain. While a lot of research has focused on the large plaques these proteins form, newer studies suggest that the real damage may begin earlier where individual AB monomers start to misfold and stick together in a process called primary nucleation (Jalalai et. al 2023). These early misfolded shapes may be key to understanding how toxic aggregates form in the first place (Arutyunyan et al 2025). This part of the aggregation pathway is still not fully understood, and I want to explore whether these early structures could be targeted by small molecules to intervene before the damage becomes irreversible.

What is the significance of this project?

Most experimental drugs targeting Alzheimer's try to break up large aggregates or use antibodies to remove toxic species from the brain, but this approach hasn't led to a reliable treatment yet, and it often misses the window where the disease actually starts. I think there's a lot of potential in looking earlier especially since small molecules could offer a more accessible treatment option. They're easier to deliver, more likely to cross the BBB, and could possibly prevent aggregation rather than just respond to it.

What are the goals of the project?

This project will use machine learning and computational chemistry to:

1. Model known AB monomer structures involved in primary nucleation.

2. Generate and screen small molecule candidates using ML-based molecular generation and docking approaches.

3. Rank and analyze high-confidence hits for their potential to interfere with these high-potential primary nucleation intermediates.

4. Produce a prioritized shortlist of compounds that could be tested in future wet-lab experiments which would form the foundation for a new therapeutic approach to Alzheimer's prevention.