Google LLC scientists Demis Hassabis and John Jumper have won this year’s Nobel Prize in chemistry together with biochemistry professor David Baker.
The trio received the award for their contributions to protein research. In an announcement today, the Royal Swedish Academy of Sciences detailed Hassabis and Jumper jointly won half the Nobel Prize for their work on AlphaFold2. This is an artificial intelligence system that Google DeepMind, the search giant’s machine learning lab, released in 2020 to advance protein research. The software solved a computational challenge that puzzled scientists for 50 years.
The behavior of proteins, the building blocks of life, is heavily influenced by their structure. Until DeepMind’s breakthrough, mapping out the structure of a protein required multimillion-dollar equipment and up to years of research. AlphaFold2 can perform the task automatically in a few minutes.
Hassabis and Jumper led the team that created AlphaFold2. Hassabis is the CEO of the AI research group while Jumper is a senior research scientist. Last year, DeepMind introduced a successor to AlphaFold2 that is more accurate and can predict the structures of not only proteins but also other biological molecules.
“The AlphaFold2 team immediately created large databases of predicted protein structures, first for the human proteome and then for the majority of sequences (> 200 million) available in the UniProt (Universal Protein Resource) database,” the Nobel Committee for Chemistry stated. “This means that almost overnight, we got access to orders of magnitude more structural information.”
Baker, the third recipient of this year’s Nobel Prize in chemistry, is a biochemistry professor at the University of Washington. He received the prize for his work on Top7, the first computationally designed protein “entirely different to all known existing proteins.”
Baker and his team created Top7 in 2003 using a program called Rosetta that they developed four years earlier. The software generates protein designs based on protein structures in the Protein Data Bank, a scientific database. Since the release of Top7, Baker and his team have used Rosetta to design numerous additional proteins that can be used as pharmaceuticals, vaccines, nanomaterials and sensors.
“In summary, the achievements of David Baker, Demis Hassabis and John Jumper in the fields of computational protein design and protein structure prediction are truly profound,” the Nobel Committee for Chemistry stated. “Their work has opened up a new era of biochemical and biological research, where we can now predict and design protein structures in ways that had not been possible before.”
On Tuesday, John Hopfield and Geoffrey Hinton won the Nobel Prize in physics for their contributions to AI research. Hopfield used methods from the field of condensed matter physics to create an early neural network. Hinton, who is also a Turing Award recipient, built on Hopfield’s work to develop one of the first deep learning model architectures.