AlphaFold AI now knows the structure of all proteins, which should contribute to many scientific advances.
At the end of 2020, Alphabet’s DeepMind division introduced its new protein folding prediction algorithm, AlphaFold, and helped solve a scientific dilemma that had plagued researchers for half a century. In the first year after the launch of the beta, half a million scientists around the world were able to access the results of the system and cite them more than 4,000 times in their research. Today, DeepMind is announcing a further increase in this access by further expanding the AlphaFold Protein Structure Database (AlphaFoldDB) from 1 million to 200 million records.
AlphaFold AI now knows the structure of all proteins
To do this, Alphabet has partnered with the European Bioinformatics Institute (EBI) at EMBL. This covers all proteins in the living kingdom—animals, plants, fungi, bacteria, etc. The results can be viewed at UniProt, Ensembl, and OpenTargets, or downloaded individually via GitHub, “for the human proteome and for the proteomes of 47 other organisms important for research.” and global health,” according to the AlphaFold website.
This should allow many scientific advances
“AlphaFold is a unique and historic achievement in the life sciences that demonstrates the power of artificial intelligence,”said Eric Topol, founder and director of the Scripps Translational Research Institute, in a press release. “Determining the three-dimensional structure of a protein used to take months or years, but now it only takes a few seconds. AlphaFold has already accelerated and made possible major discoveries, including understanding the structure of nuclear pores. And with these new building blocks spanning almost the entire protein universe, we can expect biological mysteries to be solved every day.”
AlphaFold is used in many applications, whether it is research on leprosy or Chagas disease, bee conservation, or even plastic pollution. DeepMind has also developed artificial intelligence that can perform better than human players, master games without even knowing the rules beforehand, or even improve traffic. DeepMind co-founder Mustafa Sulaiman left the company in January to start a new company, Inflection.AI, with LinkedIn co-founder Reid Hoffman.