Imagine trying to predict how this will fold. That’s what it used to be like to predict the shape of proteins. But a breakthrough changed all of that and just won the 2024 Nobel Prize in Chemistry.
David Baker, Demis Hassabis and John Jumper took home the prize for their work in computational protein design and structure prediction. If you want to build a protein, you have to know the order of its amino acids; that’s what determines its shape, and the shape determines what it does.
In 2003, Baker’s team drew a protein shape that didn’t exist in nature and predicted its amino acid sequence with a computer program. Then, they created that protein in the lab, and its structure matched their model almost perfectly.
Hassabis and Jumper worked in the other direction, using neural networks to predict shape from sequence. The AI model they created achieved almost 90% …