“New era in digital biology”: AI reveals structures of almost every known protein |  Science

“New era in digital biology”: AI reveals structures of almost every known protein | Science

What a difference a year makes. Twelve months ago, the artificial intelligence (AI) company DeepMind stunned many scientists by releasing predicted structures for about 350,000 proteins, which were acknowledged as part of the work Science‘s 2021 breakthrough of the year. Yesterday, DeepMind and its partners went much, much further. The company revealed the likely structures of almost every known protein, more than 200 million from bacteria to humans, a remarkable achievement for AI and a potential treasure trove for drug development and evolutionary studies.

“We are now releasing the structures for the entire protein universe,” said Demis Hassabis, founder and CEO of DeepMind, at a press conference in London.

The structural bounty comes from AlphaFold, one of the new AI programs that cracked the protein folding problem, the long-standing challenge of accurately inferring the 3D shapes of proteins from their amino acid sequences. AlphaFold’s newly predicted structures were yesterday added to an existing database through a partnership with the European Bioinformatics Institute (EMBL-EBI) of the European Molecular Biology Laboratory. The database “has provided structural biologists with this powerful new tool that lets you look up a protein’s 3D structure almost as easily as you can do a keyword Google search,” Hassabis said.

Eric Topol, director of the Scripps Research Translational Institute, echoed the astonishment of many outside scholars. “AlphaFold is the unique and meaningful advance in life sciences that demonstrates the power of AI,” he tweeted. “With this new addition of structures that shed light on almost the entire protein universe, we can expect more biological mysteries to be solved every day.”

The release of the DeepMind structure is “remarkable,” said Ewan Birney, Deputy Director General of EMBL, at the press conference. “It will make many researchers around the world think about what experiments they can do now.”

Proteins resolved by AlphaFold come from organisms ranging from bacteria to plants to vertebrates, including mice, zebrafish and humans. Kathryn Tunyasuvunakool, a DeepMind researcher, said it took AlphaFold about 10 to 20 seconds to make each protein prediction. The company had to work closely with EMBL-EBI, she noted, to figure out how to represent the immense number of structures in the database.

DeepMind says more than 500,000 researchers have used the database since it was launched last year. Hassabis predicted a “new era in digital biology” where drug developers could move from AI-predicted structures of proteins important to every medical condition to using AI to design small molecules that express these proteins influence – and thus treat a disease.

Others use the structural predictions to design vaccine candidates, study fundamental biological questions such as how the so-called nuclear pore complex controls which molecules enter the nucleus, or study the evolution of proteins as first life evolved.

However, Hassabis warned that declassifying the structures is only a starting point. “There’s obviously still a lot of biology and a lot of chemistry that needs to be done.”

Leave a Comment

Your email address will not be published.