University of Newcastle upon Tyne and RxCelerate Limited KTP 24_25 R4

Proteins flicker through brief, unstable shapes that current lab tools cannot detect, and a new computational method aims to catch these fleeting forms in action. Most drugs work by latching onto proteins, but they only target the stable, well-known shapes. Many proteins briefly twist into alternative configurations that could offer entirely new ways to block disease—yet these states vanish too quickly for microscopes or crystallography to capture. The researchers are building a computer workflow that predicts these invisible protein states from basic physical principles, without needing experimental data to start. If the method works, it could reveal drug targets that have been hidden inside common proteins for decades. Instead of screening millions of molecules against a static protein shape, pharmaceutical companies could design drugs that lock onto a transient state only a handful of atoms shift away from the normal form. This would open up classes of proteins currently considered "undruggable" to entirely new mechanisms of action. The project is fundamentally computational and exploratory—it tests whether prediction alone can reliably find these states. Success would not immediately produce a medicine, but it would give drug hunters a map to territory they have never been able to see.