Scientists at Google DeepMind have been awarded a $3 million prize for developing an artificial intelligence (AI) system that has predicted how almost all known proteins fold into their 3D shape.
One of this year’s Breakthrough Awards in Life Sciences went to Demis Hassabis, co-founder and CEO of DeepMind, which created the protein prediction program known as AlphaFold, and John Jumper, senior research scientist at DeepMind, the Foundation of the Breakthrough Award. Announced (opens in a new tab) Thursday (September 22).
The open source program makes its predictions based on a protein’s amino acid sequence, or the molecular units that make up the protein, Live Science previously reported. These individual units are linked together into a long chain which is then “folded” into a 3D shape. The 3D structure of a protein dictates what that protein can do, whether it’s cutting DNA or marking dangerous pathogens for destruction, so being able to infer the shape of proteins from their amino acid sequence is incredibly powerful.
The Breakthrough Awards recognize leading researchers in the fields of fundamental physics, life sciences and math. Each award comes with a $3 million prize pool, provided by founding sponsors Sergey Brin; Priscilla Chan and Mark Zuckerberg; Yuri and Julia Milner; and Anne Wojcicki.
Related: 2 Scientists Win $3M ‘Innovation Prize’ for mRNA Technology Behind COVID-19 Vaccines
“Proteins are the nanomachines that make cells work, and predicting their 3D structure from their amino acid sequence is critical to understanding how life works,” the foundation statement says. “With their team at DeepMind, Hassabis and Jumper conceived and built a deep learning system that accurately and rapidly models protein structure.”
Using AlphaFold, the DeepMind team has compiled a database of some 200 million protein structures, including proteins made by plants, bacteria, fungi, and animals, Live Science previously reported. This database includes nearly every cataloged protein known to science.
The AI system “learned” to assemble these shapes by studying known protein structures compiled from existing databases. These protein structures had been carefully visualized with a technique called X-ray crystallography, which consists of eliminating crystalline protein structures with X-rays and then measuring how those rays diffract.
Within these existing databases, AlphaFold identified patterns between the amino acid sequences of proteins and their final three-dimensional shapes. Then, using a neural network, an algorithm loosely inspired by how neurons process information in the brain — the AI used this information to iteratively improve its ability to predict protein structures, both known and unknown.
“It has been very inspiring to see the myriad ways the research community has taken up AlphaFold, using it for everything from understanding disease, protecting honey bees, cracking biological puzzles, and delving into the origins of life itself,” Hassabis wrote in a statements (opens in a new tab) published in July.
“As pioneers in the emerging field of ‘digital biology,’ we are excited to see the enormous potential of AI begin to materialize as one of humanity’s most useful tools for advancing scientific discovery and understanding the fundamental mechanisms of life,” he wrote. .
Originally published on Live Science.