Tech executives have long touted the potential for their AI models to accelerate the pace of biology discoveries. Last week, they announced one. In a September 23 announcement, Anthropic claimed that its large language model Claude had identified an enzyme system with properties “reminiscent of Crispr,” the Nobel Prize-winning gene-editing tool.
According to Anthropic, about 950 Claude agents running simultaneously found the interesting genetic sequences in 21.5 hours. “I sincerely compliment Anthropic for telling the world about their discovery,” says Fyodor Urnov, a gene-editing expert at the University of California, Berkeley and director for therapeutic R&D at its Innovative Genomics Institute. (IGI is collaborating with Anthropic, but was not involved with the new discovery.)
Other scientists are less enthusiastic about the finding, which is the first to come out of a research group formed by Anthropic earlier this year. “The experiments are still in the queue. The PR is already live,” says Le Cong, a professor at Stanford University focused on integrating AI into genome engineering research.
Anthropic, which has set up a wet lab for drug discovery, has been clear that the finding isn’t the end of its work. (“We don’t yet understand what this system does,” the company said in an X post, “but only a handful of known systems share its features, and all of them are able to cut, copy, and paste DNA.”) One physical experiment was included in a technical report, which has not been peer-reviewed. However, a lot more needs to be done to determine the significance of this discovery. It remains unknown whether this enzyme system can be used as a gene-editing tool and, if so, whether it’s a useful gene-editing tool.
“Let’s say we are on Santa Monica Beach and trying to scan through all the sand to find a diamond,” Cong says. “AI found this thing that looks very shiny, and then you have to go back to the lab to know—is it glass? Is it a diamond?”
Claude found this shiny object after researchers at Anthropic prompted it to search through huge genomic databases for “interesting new examples” of reverse transcriptases, proteins that copy RNA into DNA. This is the opposite of the normal process in our cells, in which DNA is transcribed into RNA, but is used by other organisms in a variety of biological functions, including inserting segments of DNA.
Claude agents initially identified more than 200,000 possible reverse transcriptases, according to Anthropic, and, after picking out several thousand that appeared to be new, narrowed the scope even further, eventually finding an “unusual” reverse transcriptase family containing a long region of repeat DNA sequences that resemble Crispr. Anthropic is calling this system ART, short for array-associated reverse transcriptases. The system is found in jumbo phages, large viruses that infect bacteria.
“I can see by eye a tandem repeat array … that’s a Crispr-like … repeat array?!” an AI agent wrote. The agent also acknowledged that this system could be a retron, as noted in the technical report. Crispr and retrons are both bacterial immune systems, and while retrons have some utility in gene editing, they’re not the same multi-tool Crispr is. It’s therefore not too surprising Anthropic’s blog post played up the Crispr comparison.
Seth Shipman, associate investigator at the Gladstone Institutes, doesn’t think what Anthropic found is a new Crispr system, but says the discovery is still interesting. “The novel thing is how they found it, not what it is,” he says. His lab has used retrons to make gene-editing systems, so it is possible to use retrons in this way.