The Algorithm That Solved Biology's 50-Year Grand Challenge
DeepMind's AlphaFold 2 cracked the protein folding problem with unprecedented accuracy, stunning the scientific world overnight.
Author: AlphaFold Monomer v2.0 pipeline / Public domain
This is a big deal. In some sense the problem is solved.
— John Moult
The Algorithm That Solved Biology’s 50-Year Grand Challenge (2020)
In 2020, AlphaFold 2, developed by DeepMind’s Demis Hassabis and led by researcher John Jumper, revolutionized the field of protein structure prediction by outperforming all other entries at the CASP14 competition. AlphaFold placed first overall, achieving unprecedented accuracy in predicting protein structures without relying on existing template structures. This breakthrough marked a significant milestone in the 50-year quest to solve this grand challenge in biology.
Why it matters: AlphaFold 2’s success underscores the transformative potential of AI in advancing fundamental scientific research. By accurately predicting protein structures, it has enabled significant progress in drug design, enzyme engineering, and molecular biology, highlighting AI’s role in accelerating scientific discovery beyond optimizing commercial products.
Further reading:
Why This Mattered
Protein structure prediction had been an unsolved grand challenge in biology since the 1970s. AlphaFold 2's breakthrough at CASP14 demonstrated that AI could accelerate fundamental scientific discovery, not just optimize commercial products. It has since predicted structures for nearly every known protein, transforming drug design, enzyme engineering, and molecular biology.
