The Membrane Biophysics Group

AlphaFold is an AI system developed by DeepMind that predicts the 3D structure of a protein from its amino acid sequence. It regularly achieves accuracy competitive with experiments. The information released today includes over 360000 atomic resolution structuresr; about double than the total number of experimental structures of macromolecules available to date via the PDB (of which only ~100000 correspond to unique proteins). In coming months, the AlphaFold Protein Structure Database plans to expand 'to cover a large proportion of all catalogued proteins' (over 100 million!).
In an article in the journal Structure Stephen Burley and Helen Berman name this event A cornerstone for artificial intelligence approaches to protein structure prediction. We believe that the latest achievements of AlphaFold, able to accurately predict the structures of proteins in numbers that reach the complete proteome of organisms, are only comparable with the scientific revolution in the 50's of the 20^th century, represented by the experimental elucidation of the structure of the DNA double helix (Rosalind Franklin together with Watson, Crick and Wilkins) and first structures of globular proteins (Perutz and Kendrew). The AlphaFold DB has been jointly developed by DeepMind and the EMBL's European Bioinformatics Institute (EMBL-EBI). It will undoubtedly boost our understanding of biology at a molecular level and quickly become an invaluable new tool to accelerate scientific research in all life and molecular science fields.