DeepMind researchers claim to have solved «the problem of protein folding», a task that has tormented scientists for more than 50 years
Some scientists spend their lives trying to identify the shape of tiny proteins in the human body
Proteins are the microscopic mechanisms that determine the behavior of viruses, bacteria, of the human body and all living things They start as chains of chemical compounds, before twisting and bending into three-dimensional shapes that define what they can do – and what they can't do
For biologists, identifying the precise shape of a protein often takes months, years or even decades of experimentation It requires skills, intelligence and more than a little elbow grease Sometimes, they never succeed
Now, an artificial intelligence lab in London has built a computer system that can get the job done in hours, even a few minutes
DeepMind, a lab owned by the same parent company as Google, said on Monday that his system, called AlphaFold, had solved what is called «the problem of protein folding«Given the chain of amino acids that make up a protein, the system can quickly and reliably predict its three-dimensional shape
This long-sought breakthrough could accelerate the ability to understand diseases, to develop new drugs and unravel the mysteries of the human body
Computer scientists have struggled to build such a system for more than 50 years For 25 last, they measured and compared their efforts through a global competition called the Critical Appraisal of Structure Prediction, or CUNESP Until now, no competitor had even come close to solving the problem
DeepMind solved the problem with a wide range of proteins, achieving a level of precision that rivals physical experiments Many scientists had assumed the time was still far away, even decades.
«I always hoped to live to see this day», said John Moult, professor at the University of Maryland who helped establish CUNESP in 1994 and continues to oversee the biennial competition «But it wasn't always obvious that I was going to get there”
Within the framework of CUNESP, DeepMind's technology has been reviewed by Dr Moult and other researchers overseeing the competition
If DeepMind's methods can be refined, according to him and other researchers, they could accelerate the development of new drugs as well as efforts to apply existing drugs to new viruses and diseases
Breakthrough comes too late to have a significant impact on the coronavirus But researchers believe DeepMind's methods could speed up response to future pandemics Some believe it could also help scientists better understand genetic diseases such as Alzheimer's or cystic fibrosis
However, experts have warned that this technology will affect only a small part of the long process by which scientists identify new drugs and analyze diseases. It was also unclear when or how DeepMind would share its technology with other researchers..
DeepMind is one of the key players in a radical change that has swept through universities, the tech industry and the medical community over the years 10 last years. Thanks to an artificial intelligence technology called a neural network, machines can now learn to perform many tasks that were once beyond their reach – and sometimes beyond the reach of humans.
A neural network is a mathematical system freely modeled on the neural network of the human brain It learns skills by analyzing large amounts of data Identifying patterns in thousands of photos of cats, for example, he can learn to recognize a cat
This is the technology that recognizes faces in the photos you post to Facebook, identifies the commands you are barking on your smartphone and translates one language to another on Skype and other services DeepMind uses this technology to predict the shape of proteins
If scientists can predict the shape of a protein in the human body, they can determine how other molecules will physically bind or attach themselves to it This is how drugs are developed: a drug binds to particular proteins in your body and changes their behavior
By analyzing thousands of known proteins and their physical forms, a neural network can learn to predict the shapes of others By 2018, using this method, DeepMind entered the CUNESP competition for the first time and its system outperformed all other competitors, signaling a significant change But his team of biologists, physicists and computer scientists, led by a researcher named John Jumper, was far from solving the ultimate problem.
Over the next two years, Dr Jumper and his team have designed a whole new type of neural network specifically for protein folding, which resulted in a huge leap forward in precision. Their latest version provides a powerful solution, although imperfect, the problem of protein folding, said Kathryn Tunyasuvunakool, DeepMind researcher.
The system can accurately predict the shape of a protein about two-thirds of the time, according to the results of the CUNESP contest And his errors with these proteins are smaller than the width of an atom – an error rate that rivals physical experiments
“Most atoms are found within an atomic diameter of where they are in the experimental structure”, said Dr Moult, the contest organizer “And with those who are not, there are other possible explanations for the differences”
Andrei Lupas, director of the department of protein evolution at the Max Planck Institute for Developmental Biology in Germany, is among those who have worked with AlphaFold He is part of a team that spent a decade trying to determine the physical form of a particular protein in a tiny, bacteria-like organism called an archeon.
This protein straddles the membrane of individual cells – part is inside the cell, part is outside – and that makes it difficult for scientists like Dr Lupas to determine the shape of the protein in the lab Even after a decade, he couldn't identify the shape
While these methods continue to improve, did he declare, they could be a particularly useful way to determine whether a new virus can be treated with a cocktail of existing drugs
“We could start testing all compounds approved for use in humans”, Dr Lupas said «We could face the next pandemic with the drugs we already have”
During the current pandemic, a simpler form of artificial intelligence has proved useful in some cases A system developed by another London company, BenevolentAI, helped identify an existing drug, baricitinib, that could be used to treat critically ill Covid-19 patients. Researchers have now completed a clinical trial, although the results have not yet been published
As researchers continue to improve technology, AlphaFold could further accelerate this type of drug reuse, as well as the development of completely new vaccines, especially if we encounter a virus even less understood than Covid-19
David Baker, director of the Institute for Protein Design at the University of Washington, who uses similar computer technology to design anti-coronavirus drugs, stated that DeepMind's methods could speed up this work.
«We were able to design proteins that neutralize coronaviruses in several months», did he declare «But our goal is to do this sort of thing in a few weeks”
However, the speed of development has to face other problems, like massive clinical trials, said Dr Vincent Marconi, researcher at Emory University of Atlanta who helped lead the baricitinib trial «This takes time», he says
But DeepMind's methods could be a way to determine if a clinical trial will fail due to toxic reactions or other issues, at least in some cases
Demis Hassabis, CEO and co-founder of DeepMind, said the company plans to release details describing its work, but that probably wouldn't happen until next year. He also said the company was exploring ways to share the technology itself with other scientists..
DeepMind is a research lab It does not sell products directly to other labs or companies But it could work with other companies to share access to its technology over the Internet
The lab's greatest advances in the past have been in games He built systems that outperformed human performance on the old strategy game Go and the popular StarCraft video game – extremely technical achievements without practical application Now, the DeepMind team is eager to push their artificial intelligence technology into the real world
«We don't want to be a leading company», Dr Jumper said «We want real biological relevance”
DeepMind, Artificial intelligence, Protein, Biology, CASP, Folding proteins
News – US – London Aje Lab claims breakthrough that could accelerate drug discovery
Associated title :
– The London AI Lab Says Breakthrough That Could Accelerate Drug Discovery
– & # 39; The game has changed& # 39; AI triumphs in the resolution of protein structures
– In a major scientific advance, AI predicts the exact form of protein
– DeepMind solves the & # 39 old of 50 years ; big challenge & # 39; with protein folding AI
– DeepMind AI makes revolutionary scientific breakthrough
– Alphabet& # 39; s DeepMind takes historic milestone in AI-based protein structure prediction
– DeepMind's protein folding AI& # 39; solved a period of 50 years- old big challenge of biology
– DeepMind AI Solution at 50 years -The Old Science Challenge could «revolutionize medical research»
– How do proteins fold??
– DeepMind's enhanced artificial intelligence& for predicting protein folding could accelerate drug discovery