The company, DeepMind, owned by Alphabet, has developed an artificial intelligence program that can accurately predict the form that proteins will deviate to within days, solving a 50-year-old “big challenge” that could pave the way for a better understanding of diseases and detection. Medicines, according to a CNBC report.
Every living cell contains thousands of different proteins that keep them alive and healthy. Predicting what a protein will look like is important because it determines its function, and nearly all diseases, including cancer and dementia, are related to how the proteins work.
Professor Dame Janet Thornton of the European Institute of Bioinformatics told reporters over the phone, “Proteins are the most beautiful shapes, and the ability to predict exactly which way you bend is really very difficult and has occupied a lot of people for many years.”
The British Research Laboratory’s AI system, AlphaFold, entered into a competition organized by a group called CASP, which is an acronym for Critical Evaluation of Structure Prediction. It is a community experiment organization whose mission is to accelerate solutions to one problem: how to calculate the 3D structure of protein molecules.
CASEB has been monitoring the field for 25 years, and it said on Monday that DeepMind’s AlphaFold system achieved unparalleled levels of accuracy in predicting protein structure.
“DeepMind has leaped forward,” said Professor John Mollet, Head of CASP, in a press call before the announcement. “The great 50-year-old challenge in computer science has largely been solved.”
Molt added that there are “little big implications for the line of drug design” and in the newly emerging field of protein design.
With nearly a thousand employees and along with no revenue, DeepMind has become a costumed company to parent company Alphabet, yet it has emerged as one of the leaders in the global AI race.
Google CEO Sundar Pichai welcomed the hack on Twitter.
.@DeepMind‘s incredible AI-powered protein folding breakthrough will help us better understand one of life’s fundamental building blocks + enable researchers to tackle new and hard problems, from fighting diseases to environmental sustainability. https://t.co/kpr8EAx34h
– Sundar Pichai (undsundarpichai) November 30, 2020
“The ultimate vision behind DeepMind was to build a general artificial intelligence, and then use it to help us better understand the world around us by dramatically accelerating the pace of scientific discovery,” said Demis Hasabis, co-founder and CEO of DeepMind.
The company, which Google bought for $ 600 million in 2014, is best known for creating artificial intelligence systems that can play games like Space Invaders and an old Chinese game called Go. However, she has always said that she wants to make a greater scientific impact.
“Games are a great proof ground for efficiently developing and testing generic algorithms, which we once hoped would take them to real-world scales like scientific problems,” said Hasabis, adding, “We feel that Alphafold is the first proof of this thesis. These algorithms are now mature enough and robust enough.” Enough to be applicable to really difficult scientific problems. “
DeepMind was also entered in the CASEB Protein Prediction Competition in 2018. His results at the time were impressive. “The team knew it was far from producing something really strong biologically related or competing with the experiment,” said John Gamper, AlphaFold’s team leader at DeepMind.
But their efforts seem to have paid off. “We really believe we have built a system that provides valid and actionable information to experimental biologists,” he said.