
Discover more from AI in Business
DeepMind’s AlphaFold Release Launches Era of ‘Digital Biology’ Powered by AI
Release of database of 200 million protein structures is seen as a major advance, opening up new opportunities for research around sustainability, fuel, food and diseases
By John P. Desmond, Editor, AI in Busines

DeepMind’s announcement that its AlphaFold tool has successfully predicted the structure of nearly all proteins known to science, and its database of 200 million proteins is being offered for free, is a major advance in the contribution of AI to science.
Google/Alphabet’s DeepMind unit, based in London, introduced AlphaFold in 2020, instantly advancing the understanding of how proteins are structured, work that scientists had spent decades trying to understand. DeepMind last year released the structures of one million proteins, in its AlphaFold Protein Structure Database.
The new release is a major advance—increasing the number of protein structures included from one million to 200 million—that includes structures for “plants, bacteria, animals, and many, many other organisms, opening up huge opportunities for AlphaFold to have impact on important issues such as sustainability, fuel, food insecurity, and neglected diseases,” stated Demis Hassabis, DeepMind’s founder and CEO, on a call as reported in MIT Technology Review

DeepMind’s announcement was well-received by scientists, who see it as helping to speed innovation in drug discovery and biology. “AlphaFold is probably the most major contribution from the AI community to the scientific community,” stated Jian Peng, a computer science professor at the University of Illinois Urbana-Champaign who specializes in computational biology, to MTR.
Hassabis described a push into “digital biology” where “AI and computational methods can help to understand and model important biological processes.” Hassabis also heads a new venture of Alphabet’s called Isomorphic Labs, which is developing AI for drug discovery.
AlphaFold’s protein structures were released into an existing database through a partnership with the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI). The release is being well-received in the scientific community.
“AlphaFold is the singular and momentous advance in life science that demonstrates the power of AI,” tweeted Eric Topol, director of the Scripps Research Translational Institute, according to an account in Science. “With this new addition of structures illuminating nearly the entire protein universe, we can expect more biological mysteries to be solved each day.”
Determining the 3D structure of a protein used to take many months or years, and “it now takes seconds,” Topol stated in a tweet, adding, “Alpha Fold has already accelerated and enabled massive discoveries.”
Ewan Birney, deputy director general of EMBL, stated at the press conference that AlphaFold “will make many researchers around the world think about what experiments they can now do.”
Kathryn Tunyasuvunakool, a DeepMind research scientist also at the press conference and also quoted in Science, stated that it took AlphaFold roughly 10 to 20 seconds to make each protein prediction. The company had to work closely with EMBL-EBI, she noted, to determine how best to present the huge number to figure out how to present the vast number of structures in the database
Hassabis predicted a “new era in digital biology that could, for example, encompass using AI to design small molecules to treat an illness. He suggested this new era is in an early stage. With “still obviously a lot of biology and a lot of chemistry that has to be done.”
In a release on the blog of DeepMind, Hassabis said the achievement of his team so far has proved “our founding thesis: that artificial intelligence can dramatically accelerate scientific discovery and in turn advance humanity.”
Potential Risks Not Detailed, Yet
In considering how to release the AlphaFold models to the world, Hassabis said DeepMind sought input from more than 30 experts across biology research, security, ethics and safety, “to help us understand how to share the benefits of AlhaFold … in a way that would maximize potential benefit and minimize potential risk.” He did not elaborate on the potential risks.
He listed in the blog post a number of AlphaFold predictions that have been referenced in publications. These included: a candidate protein for inclusion in a vaccine; the regulation of a cellular process; a bacterial protein that can trigger ice formation at relatively high temperatures; a plant protein representing a potential new structural superfamily unlike anything seen before; and a protein involved in the immune system of egg-laying animals including honeybees.
“We’ve been amazed by the rate at which AlphaFold has already become an essential tool for hundreds of thousands of scientists in labs and universities across the world to help them in their important work,” Hassabis stated.
The stage is set for much more biology research incorporating AI to take place. “AlphaFold is a glimpse of the future, and what might be possible with computational and AI methods applied to biology,” Hassabis stated, adding, “Just as maths is the perfect description language for physics, we believe AI might turn out to be just the right technique to cope with the dynamic complexity of biology.”
A recent account from SciTechDaily expands on the impact of the new AlphaFold release. “Many other AI research organizations have now entered the field and are building on AlphaFold’s advances to create further breakthroughs. This is truly a new era in structural biology, and AI-based methods are going to drive incredible progress,” stated John Jumper, Research Scientist and AlphaFold Lead at DeepMind.
Sameer Velankar, Team Leader at EMBL-EBI’s Protein Data Bank in Europe, stated, “AlphaFold has sent ripples through the molecular biology community. In the past year alone, there have been over a thousand scientific articles on a broad range of research topics which use AlphaFold structures; I have never seen anything like it.” and now the number of protein structure predicts has been multiplied dramatically. “Imagine the impact of having over 200 million protein structure predictions openly accessible in the AlphaFold database.”
Read the source articles and information in MIT Technology Review, in Science, in the blog of DeepMind and in SciTechDaily.
(Write to the editor here.)