NEW YORK — The 2024 Nobel Prize in Chemistry was awarded to a trio of researchers that pioneered computation approaches to designing proteins and predicting their structures, the Royal Swedish Academy of Sciences announced Wednesday.
The prize is shared by Demis Hassabis and John Jumper at Google DeepMind for their work building an AI tool for predicting protein structure and University of Washington researcher David Baker for his research into computational protein design.
Hassabis and Jumper are the developers of AlphaFold2, an AI model that predicts protein structure based on amino acid sequence. Hassabis is the CEO and founder of DeepMind, which Google acquired in 2014. Jumper joined the company in 2017, working with Hassabis to improve the company's existing protein prediction model, AlphaFold, ultimately boosting the accuracy of its predictions from around 60 percent to greater than 90 percent and providing a tool for rapidly generating protein structural data.
This model, AlphaFold2, has since become widely used in drug discovery and development and in protein research, including applications within proteomics like protein-protein interaction experiments. This year, DeepMind released a new version called AlphaFold3, which expands beyond the prediction of protein structure and protein-protein interactions to enable predictions of a variety of biomolecular structures, including protein-nucleic acid complexes, protein-small molecule complexes, and protein post-translational modifications.
Baker also began his career investigating the mechanisms behind protein folding and likewise developed a computer program, Rosetta, for predicting protein structures. Finding success in this area, he moved to tackle the question from the opposite angle. According to the Royal Swedish Academy of Sciences, he realized that the program's ability to predict structure from amino acids sequence suggested it might be possible to reverse the process — to start with a desired protein structure and predict the amino acid sequence that would produce that structure.
While scientists had previously altered existing proteins to adapt them to various purposes, Baker set out to create entirely new proteins. Using Rosetta, he identified an amino acid sequence that would produce a desired protein structure then produced a gene that would code for that amino acid sequence. He expressed the gene and protein in a bacterial model and analyzed the resulting structure using X-ray crystallography, finding that the protein they produced, named Top7, closely matched the intended structure. Published in 2003, the work marked a dramatic step forward for the field of synthetic biology. Since then his lab has produced a number of novel proteins with applications in areas such as biosensing and drug development.