Artificial intelligence has solved one of the biggest puzzles in biology, predicting the shape of every protein expressed in the human body.
The research was conducted by London-based company AI DeepMind, which used its AlphaFold algorithm to build the most complete and accurate database of the human protein, which supports human health and disease.
Last week, DeepMind published the methods and code for its model, AlphaFold2 in Nature, showing that it could predict known protein structures with near-perfect accuracy.
The follow-up with his second paper Nature in Many Weeks, published Thursday, demonstrates that the model could reliably predict the structural position for nearly 60 percent of amino acids, protective blocks, in the human body, and even in a mass. of other organisms such as the fruit fly, the mouse and the E.coli battery.
The structural situation for only about 30 percent of amino acids was known before. Understanding the position of amino acids allows researchers to predict the three-dimensional structure of a protein.
The set of 350,000 protein structure predictions is now available through a public database hosted by the European Institute of Bioinformatics in the European Laboratory of Molecular Biology (EMBL-EBI).
Edith Heard, director general of the EMBL Department. “Applications are limited only by our imagination.”
Protein structures matter because they tell you how proteins do their job. Knowing the shape of a protein – say a Y-shaped antibody – tells scientists more about what the role of this protein is.
Deformed proteins can cause diseases such as Alzheimer’s, Parkinson’s and cystic fibrosis. Being able to easily predict the shape of a protein could allow scientists to control and modify it, to improve its function by changing its DNA sequence, or targeting drugs that could attack it.
Accurate prediction of the structure of a protein from its DNA sequence has been one of the greatest challenges of biology. Current experimental methods for determining the shape of a single protein take months or years in a laboratory, which is why only about 180,000 protein structures have been resolved, of the more than 200m proteins known to living things.
“We believe this will represent the most significant contribution that AI has made to advancing the state of scientific knowledge to date,” said Demis Hassabis, CEO of DeepMind. “Our ambitions are to expand.” [the database] in the coming months across the universe of proteins over 200m of protein. ”
Scientists who did not participate in DeepMind’s research used phrases such as “spinal cord” and “transformative” to describe the impact of the advance, comparing the data set to the human genome.
“It was one of those moments when my hair stood on end around my neck,” said John McGeehan, director of the Center for Enzyme Innovation at the University of Portsmouth, and a structural biologist who has tested the AlphaFold algorithm over the last few months.
“We are able to use this information directly to develop faster enzymes to break down plastics. Those experiments are underway immediately, so the acceleration to that project here is several years.”
AlphaFold is not without limitations. Proteins are dynamic molecules that change shape constantly depending on what they bind to, but the DeepMind algorithm can only predict the static structure of a protein, said Minkyung Baek, a researcher at the Institute for Protein Design of the University of Washington.
However, his biggest contribution to scientists has been the fact that he was open, he said. “They protested last year [this] it’s all possible but he didn’t provide any code, so people knew he was there but couldn’t use it. ”
In the seven months following DeepMind’s announcement, Baek and colleagues used DeepMind’s idea to build their open version of the algorithm they called RosettaFold, and it was published in the journal. Science last week. “I am really happy to have made everything publicly available, it is a huge contribution to biological research and also to commercial pharma,” he said. “Now more people can benefit from his method.” [and] it advances the field much more rapidly. ”