How Google’s AI solved grand biology challenge to win chemistry Nobel

The 2024 Nobel Prize in Chemistry has been awarded to a team of scientists for their groundbreaking work in using artificial intelligence to predict protein structures, revolutionising the field of structural biology.
The Royal Swedish Academy of Sciences awarded one-half of the 2024 Nobel Prize in Chemistry to David Baker “for computational protein design” and the other half jointly to Demis Hassabis and John M. Jumper from Google’s DeepMind “for protein structure prediction.”
While Demis Hassabis is CEO of Google DeepMind in London, John M. Jumper is a Senior Research Scientist at DeepMind.
Proteins are large, complex molecules that play many critical roles in the body. They are made up of smaller units called amino acids, which are linked together in long chains. There are 20 different types of amino acids, and the specific sequence in which they are arranged determines the protein’s structure and function.
The team, led by researchers from Google DeepMind and their academic collaborators, developed AlphaFold, an AI system that has transformed our understanding of protein structures and their functions.

AlphaFold Protein Structure Database has grown to include over 200 million protein structures. (Photo: Google DeepMind)

AlphaFold, first introduced in 2020, demonstrated an unprecedented ability to predict protein structures with atomic-level accuracy.
This achievement solved a 50-year-old grand challenge in biology, known as the “protein folding problem,” which had long stumped scientists due to the astronomical number of possible configurations for even a single protein.
The Nobel Committee praised the team’s work for its far-reaching implications across various scientific disciplines. By accurately predicting protein structures, AlphaFold has accelerated research in areas such as drug discovery, disease understanding, and environmental science.

One of the key innovations behind AlphaFold was its use of deep learning techniques, including a novel neural network architecture called Evoformer.
This approach allowed the system to leverage evolutionary information from multiple sequence alignments and incorporate physical and biological constraints into its predictions.
The impact of AlphaFold has been profound and rapid. Since its public release, the AlphaFold Protein Structure Database has grown to include over 200 million protein structures, covering nearly all catalogued proteins known to science.
This vast resource has been accessed by more than half a million researchers worldwide, accelerating progress on critical issues ranging from antibiotic resistance to plastic pollution.
The Nobel laureates’ work has not only provided a powerful tool for scientific research but has also demonstrated the potential of AI to solve complex scientific problems.
“Life could not exist without proteins. That we can now predict protein structures and design our own proteins confers the greatest benefit to humankind,” the Nobel Prize announcement read.