Open Science Spotlight: AlphaFold
This week’s open science spotlight is on a story many of us already love, because it never stops being inspiring: AlphaFold.
When Google’s DeepMind’s AlphaFold2 cracked a 50-year-old grand challenge in biology – predicting a protein’s 3D structure from its amino acid sequence – the real breakthrough wasn’t just the algorithm. It was the choice of what to do with it. DeepMind open-sourced the code and, together with EMBL-EBI, built the AlphaFold Protein Structure Database, making predicted structures for virtually all 200 million proteins known to science freely available to anyone, anywhere. The ripple effects have been extraordinary. AlphaFold itself was only possible because it trained on decades of open data – most importantly the Protein Data Bank, where structural biologists have freely deposited experimentally solved protein structures since 1971. By October 2024, AlphaFold2 had been used by more than two million people across 190 countries, and work that once took years of painstaking lab work can now be done in minutes. Then it gave that gift back: DeepMind open-sourced the code and, with EMBL-EBI, released predicted structures for virtually all 200 million proteins known to science, freely, to anyone. In 2024, that arc was recognized with the ultimate stamp of approval: the Nobel Prize in Chemistry, awarded jointly to Demis Hassabis and John Jumper “for protein structure prediction,” alongside David Baker “for computational protein design”. That’s the case for open science in one story: share the data, share the tools, and watch the whole field move faster together.
Science belongs to us all – let’s keep building it that way.
By Teresa Simoes Steyn