‘A new era in digital biology’: AI reveals the structure of almost all known proteins Science

What a difference a year makes. Twelve months ago, artificial intelligence (AI) company DeepMind surprised many scientists by releasing the predicted structures for about 350,000 proteins, which it said was part of the work. ScienceSuccess of the year 2021. Yesterday, DeepMind and its partners went much, much further. The company unveiled the possible structures of nearly all known proteins, 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 structures for the entire protein universe,” DeepMind founder and CEO Demis Hassabis said at a press conference in London.

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

Eric Topol, director of the Scripps Research Translational Institute, echoed the surprise of many outside scientists. “AlphaFold is a singular and important advance in the life sciences that demonstrates the power of AI,” he tweeted. “With this new addition of structures illuminating almost the entire protein universe, we can expect more biological mysteries to be solved every day.”

The DeepMind structure release is “remarkable,” Evan Burney, EMBL’s deputy director general, said at a press conference. “It has many researchers around the world thinking about what they can do next.”

Proteins resolved by Alphafold come from organisms ranging from bacteria to plants to vertebrates including mice, zebrafish and humans. Katherine Tunyasuvunakul, a DeepMind research scientist, said it took Alphafold about 10 to 20 seconds to predict each protein. The company had to work closely with EMBL-EBI, she noted, to figure out how to introduce the huge number of structures into the database.

DeepMind says more than 500,000 researchers have used the database since its launch last year. Hasbis predicts a “new era in digital biology” in which drug developers can use AI to design small molecules that affect those proteins from AI-predicted structures of proteins important for any medical condition—and therefore treat the patient.

Others are using structure predictions to develop vaccine candidates, investigating fundamental biology questions such as the so-called nuclear pore complex gatekeeping which molecules enter the cell nucleus, or examining the evolution of proteins when life first evolved.

Hassabis, however, cautioned that the release of structures is only a starting point. “There’s still obviously a lot of biology, and a lot of chemistry, that has to be done.”

Leave a Comment

Your email address will not be published.