IIn 2018, Francis Arnold said in his Nobel Prize Lecture in Chemistry that scientists had reached the point where they could read, write, and edit any sequence of DNA. But constructing entire genes, or even entire genomes, from scratch was something only evolution could do.
Years later, shortly after helping start Ark Institute, a nonprofit research center in the Bay Area, molecular engineer Patrick Hsu discovered that it was possible to mimic the evolutionary forces Arnold was referring to. I wondered if it was. After all, DNA is a language, and with all the advances in generative AI (chatbots that can have eerily believable conversations when trained on enough text) It may not be too late to reproduce it.
In collaboration with Brian Hie, a computational biologist at Stanford University and a member of the Ark Institute, Hsu, who is also an assistant professor at the University of California, Berkeley, is working with scientists to train AI models on vast amounts of biological material. We started forming a team. Data — 300 billion DNA letters, including long sequences from his 80,000 genomes of bacteria and archaea.
This article is exclusive to STAT+ subscribers
Unlock this article to get access to even more in-depth analysis, newsletters, premium events, and networking platforms.
Do you already have an account?Login
Do you already have an account?Login
See all plans