British scientists have used artificial intelligence (AI) to create a database of “the entire protein universe” which could help pave the way for a cure for Parkinson’s, as well as other scientific breakthroughs.
DeepMind, a company based in London and owned by Alphabet, the Silicon Valley tech giant that owns Googlefirst revealed its AlphaFold technology last year when it contained the 3D shape and structure of around 350,000 proteins.
Now, the database has expanded to more than 200 million proteins, which is almost all the proteins known to science and covers animals, proteins, bacteria and everything in between.
By accurately unpicking how proteins look in reality from their genetic code, scientists can predict how they will interact with other proteins, antibodies and drugs.
“When we launched last July it was recognized as a pretty big leap forward for biology,” Dr Demis Hassabis, CEO and Founder of DeepMind, said.
“It was also a great demonstration of how AI can be used to advance scientific discovery and it has provided structural biologists with this powerful new tool to look up the 3D structure of a protein almost as easily as doing a keyword Google search and, as we know, the 3D structure of proteins is vital for understanding their function.
“You can think of [AlphaFold] as covering the entire protein universe.”
He added that more than half a million scientists have already used the ever-expanding directory for myriad projects.
‘An entire universe in just one click’
Professor Ewan Birney, Deputy Director General of the European Molecular Biology Laboratory (EMBL), which helped develop AlphaFold with DeepMind, added it was “remarkable” that the entire protein universe is now available in just one or two clicks.
“It will make many researchers around the world think about what experiments they could do,” he said.
AlphaFold has been touted as one of the biggest scientific advances in recent years and has the potential to revolutionize various fields, including pharmacology, agronomy and vaccine manufacture. It is also being used to tackle the existential issues of plastic pollution and antibiotic resistance.
One major avenue which is being aided by AlphaFold’s technological insight is that of disease treatment.
Scientists at Duke University and the National University of Singapore published a paper in May based on results made possible by AlphaFold.
They focused on a protein called phosphoprotein 1 (STIP1) which they were able to identify as something which helps protect the brain but is targeted and destroyed by the immune system of Parkinson’s patients.
Before now the structure of this protein was never known in detail and AlphaFold was able to accurately predict the structure from the DNA and amino acid arrangement of the protein.
This, the researchers said, was “in excellent agreement” with the more blurry picture of the protein made via experiments.
By enabling researchers to get over their technological and laboratory hurdles, AlphaFold has elevated almost all fields of biology. It is a watershed moment for the life sciences and will reap rewards in scores of different fields in the years to come.
‘AlphaFold was transformational’
Prof Matt Higgins, Professor of Molecular Parasitology at the University of Oxford, is using AlphaFold to fight malaria by creating new vaccines against the mosquito-borne parasite.
His protein of interest had always been “fuzzy”, despite years of lab work, he said.
“The use of AlphaFold was really transformational, giving us a really sharp view of this maaria surface protein,” Prof Higgins said.
This has now led to new vaccine candidates being produced, and they are now in testing, Prof Higgins said.
‘I’ve never seen anything like it’
“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,” said Dr Sameer Velankar, Team Leader at EMBL-EBI’s Protein Data Bank in Europe.
“And this is just the impact of one million predictions; imagine the impact of having over 200 million protein structure predictions openly accessible in the AlphaFold Database.”