AMI‑CryoML
Integrating cryo‑EM and ML to decode amyloid fibril polymorphism.
🧠 AMI‑CryoML: unraveling amyloid polymorphism
Cracking the code of amyloid polymorphism: integrating cryoEM and machine learning to unravel impact of small aggregation modulators on amyloid fibril polymorphism
PI: Jarek
Grant ID: 101244706
Start Date: 1 Feb 2026
Duration: 24 months
EU Contribution: €181,136.16
Neurodegenerative disorders such as Alzheimer’s, Parkinson’s, and prion diseases are marked by protein misfolding into amyloid fibrils. The structural diversity (polymorphism) of these fibrils critically affects their pathology, yet remains largely unexplored. AMI‑CryoML bridges cryo‑EM and machine learning to probe how small molecule modulators influence aggregation kinetics and amyloid structures.
📌 Project highlights
- 🔍 Systematic collection of aggregation modulators: chemical and mechanistic profiling
- 🧪 Thioflavin‑T assays to measure kinetic effects across proteins (prions, CsgA, α‑synuclein)
- 🧬 Cryo‑EM to resolve fibril polymorphs under varied conditions
- 💾 Extension of the AmyloGraph database to store modulator‑kinetics‑structure data
- 🤖 Machine learning models to predict modulator effects