In biology circles, Google’s AI research company DeepMind is best known for folding proteins. Its deep learning model AlphaFold, which predicts the structure of a protein from its sequence of amino acids, shared the Nobel Prize in chemistry last year with the University of Washington’s David Baker.
But DeepMind has lately been working on a much gnarlier precursor problem: predicting how DNA encodes gene regulation. In a preprint and blog post published on Wednesday, it announced its latest stab at the challenge with a new model called AlphaGenome, an early step toward possible applications in therapeutic development.
Compared with the well-defined problem of protein structure prediction, “genomics is more of a fuzzy field,” said DeepMind research engineer Natasha Latysheva. “There’s no single metric of success.” So DeepMind has gone after as many of them as it can.
STAT+ Exclusive Story
Already have an account? Log in
This article is exclusive to STAT+ subscribers
Unlock this article — plus in-depth analysis, newsletters, premium events, and news alerts.
Already have an account? Log in
Individual plans
Group plans
To read the rest of this story subscribe to STAT+.