Google’s DeepMind artificial intelligence has revealed the 3-D structure of 200 million proteins found in every living organism, giving scientists instant access to detailed information about the building blocks of life.
Scientists used to spend many months or years understanding the structure of proteins before the AI program known as AlphaFold solved one of biology’s toughest problems in November 2020. Researchers often used tools like X-rays, but complex information is now available. as fast as a Google search.
AlphaFold can predict the structure of almost all proteins, whether in animals, plants, humans, bacteria, or other organisms, known to science. This ability to quickly see the structure of a protein in three dimensions is valuable to scientists looking to cure diseases and researchers looking to solve problems like plastic pollution because it gives them a completely detailed view of the proteins that hold all the biological processes.
However, as with anything involving AI, this work comes with risks. A study in Nature in March showed that scientists’ drug discovery algorithm could, with a few small tweaks, generate toxic concoctions like the nerve agent VX and other chemical warfare molecules.
Scroll down to watch the video
Google’s DeepMind artificial intelligence can now reveal the structure of 200 million proteins that are the building blocks of life, allowing scientists to use this knowledge in many ways. Pictured above is the F2OH23.2 protein, a plant protein that represents a potential new structural superfamily unlike anything seen before.
Being able to understand the shape of a protein allows researchers to determine how it works in the body and, for example, how effective certain drugs can be when they interact with it. Pictured above: Vitellogenin, a protein that is involved in the immune system of egg-laying animals, including bees.
“It has been very inspiring to see the myriad of ways the research community has taken AlphaFold, using it for everything from understanding disease, to protecting bees, to deciphering biological puzzles, to delving into the origins of life itself,” he said. say DeepMind founder and CEO Demis Hassabis
DeepMind says it has followed a “responsible” path by consulting with more than 30 experts in biology, ethics, safety and security so that the benefits of AI could be shared in a way that minimizes potential risks.
Still, numerous experts, including former Google AI ethics staffer Timnit Gebru, have raised concerns about the technology’s potential to reinforce a concentration of power that results in discrimination.
To date, more than 500,000 researchers worldwide from 190 countries have used the AlphaFold database to view more than 2 million structures.
“It has been very inspiring to see the myriad of ways the research community has taken AlphaFold, using it for everything from understanding disease, to protecting bees, to deciphering biological puzzles, to delving into the origins of life itself,” he said. said DeepMind founder and CEO Demis Hassabis in a statement on Thursday.
So far, more than 500,000 researchers worldwide from 190 countries have used the AlphaFold database, which includes the predicted structures of plants, animals, bacteria, fungi and other organisms, to view more than 2 million structures (like the one seen above)
The new database could help everyone from scientists working on malaria vaccines to experts trying to solve ocean plastic pollution to researchers studying bee immunity. Pictured above is subunit 9 of the CCR4-NOT transcription complex, a protein that regulates an important cellular process in humans.
As with anything involving AI, this work has risks. A study in Nature in March showed that scientists’ drug discovery algorithm could, with a few small tweaks, generate toxic concoctions like the nerve agent VX. Pictured above: Nuclear pore complex protein Nup205, which is part of a large complex that acts as a gate in and out of the cell nucleus
The breakthrough announced is a collaboration between UK-based Alphabet subsidiary DeepMind and the European Bioinformatics Institute.
A TIMELINE OF GOOGLE’S DEEP AI ADVANCES
DeepMind was founded in London in 2010 and was acquired by Alphabet in 2014. It has offices in Edmonton and Montreal, Canada, Paris and Mountain View, California.
In March 2016, DeepMind’s AlphaGo program defeated Go player Lee Sedol in a challenge in Seoul. This moment proved that AI techniques were strong enough to tackle bigger problems like protein folding.
In December 2018, AlphaFold did its first public test of its performance and was ranked first. The results were published in Nature and the team is expanding.
In November 2020, AlphaFold was recognized for finding a solution to the 50-year-old “protein folding problem” that scientists had long struggled with by winning a competition and successfully predicting protein structures down to the “atomic precision”.
On July 15, 2021, Nature published AlphaFold’s detailed methodology in the paper titled “Highly accurate protein structure prediction with AlphaFold,” and DeepMind is opening the code along with 60 pages of additional information detailing the entire system.
A week later, Nature published another paper containing the predictions for the entire structure of the human proteome, that is, all the proteins in the human body. Next, DeepMind releases the AlphaFold protein structure database to give the scientific community free access to more than 350,000 total protein structures.
DeepMind adds 400,000 additional protein structures to the AlphaFold protein structure database, doubling its size.
As of January 2022, more than 300,000 researchers worldwide have used the database.
On July 28, DeepMind expands the AlphaFold protein structure database from nearly 1 million to more than 200 million structures.
“AlphaFold is the singular and momentous breakthrough in life sciences that demonstrates the power of AI. Determining the 3D structure of a protein used to take many months or years, now it takes seconds,” said Eric Topol, founder and director of the Scripps Research Translation Institute, in a statement.
“AlphaFold has already accelerated and enabled massive discoveries, including breaking the structure of the nuclear pore complex,” he explained. “And with this new addition of structures illuminating almost the entire protein universe, we can expect more biological mysteries to be solved every day.”
Finding a new drug that can cure a disease is very difficult, especially when trying to determine how the different receptors (proteins that bind to the drug) work in a family. That is, scientists often want to find a drug or molecule that targets one member of that family, but not all. This is where AlphaFold’s capabilities can help.
“This could accelerate drug discovery in a way we’ve never seen before,” said Karen Akinsanya, president of research and development at Schrodinger in New York.
Plastic pollution – the world produces around 400 million tonnes a year – is a perhaps surprising area where AlphaFold could play a key role. Scientists at the University of Portsmouth’s Enzyme Innovation Center are developing a unique solution, something they call a “fully circular plastic economy” that would use enzymes, which are proteins that speed up metabolism, to break down plastic polymers. In this way, they could be 100% recycled back to their original state, or even recycled into material that has the quality of virgin plastic, instead of polluting the oceans.
“One year after opening their protein structure database AlphaFold, DeepMind and EMBL’s European Bioinformatics Institute are expanding it from nearly 1 million to more than 200 million protein structures, covering almost every organism that has had its genome sequenced, a huge milestone,” Sundar, CEO of Alphabet. Pichai said on Twitter.
“As pioneers in the emerging field of ‘digital biology,’ we are excited to see the enormous potential of AI begin to be realized as one of humanity’s most useful tools for advancing scientific discovery and understand the fundamental mechanisms of life,” Hassabis said. .
While DeepMind prides itself on keeping the database open source to allow broad access to scientists, the company’s founder admitted during the announcement to reporters that they may have to restrict access in the future.
‘Future [systems]if they have risks, the whole community should consider different ways to give access to that system, not necessarily open source everything, because that could allow bad actors,” Hassabis said.
“Open source is not some kind of panacea,” Hassabis added. ‘It’s great when you can do that. But there are often cases where the risks can be too great.”
Alphabet CEO Sundar Pichai announced the announcement of the expanded database capabilities in a statement posted on Twitter (seen above)
Today’s announcement shows that thanks to AlphaFold, researchers have access to a much larger number of protein structures for their work
AI work is not without its risks, as experts fear that these types of programs can be misused. The latest update includes predicted structures for animals, plants, bacteria, fungi and more (as seen above)