Every two years since 1994, scientists have come together for an experiment — better described as a competition — where teams of computational biologists try to predict what proteins will look like in 3D given their amino acid sequence. In the first 24 years, none of the entrants scored better than around 40 out of 100 for the most difficult targets. Then in 2018, a team from DeepMind, a machine-learning company owned by Alphabet, entered the competition for the first time and astonishingly notched about 60 using an artificial intelligence platform called AlphaFold. In 2020, its successor, AlphaFold2, achieved near-perfect scores.
In the meeting’s closing remarks, the organizers declared the protein-structure prediction problem largely “solved.” Everyone could pack up, go home, and study another problem.
That launched one of the most-hyped scientific narratives in recent memory — much of it focused on the AI platform’s potential to revolutionize and accelerate drug development. After all, AlphaFold had reduced months or years of laborious work using X-ray crystallography to determine a single protein’s structure to minutes or seconds.
This article is exclusive to STAT+ subscribers
Unlock this article — plus daily coverage and analysis of the biotech sector — by subscribing to STAT+.
Already have an account? Log in
Already have an account? Log in
About the Author Reprints
To submit a correction request, please visit our Contact Us page.
STAT encourages you to share your voice. We welcome your commentary, criticism, and expertise on our subscriber-only platform, STAT+ Connect