Skip to content

Choose which optional content this site may load. You can change this at any time from the Cookie Settings link at the bottom of the page.

Necessary

Remembers your cookie and accessibility choices in your browser. Always on.

Allows audience measurement tools to load, if this site uses them.

Allows content from other services, such as embedded social media posts, to load. These services may set their own cookies.

Accessibility

Text size 100%

Settings are saved in this browser.

Health & Science

Anthropic Says Claude Identified a New Enzyme System in Phage DNA

Abstract blue line illustration used for IsraelNews.io Health & Science coverage

Anthropic published early results on September 23 from a biology project in which Claude agents searched large DNA sequence databases and identified a previously uncharacterized enzyme system associated with a DNA repeat array in viruses that infect bacteria.

The company calls the system ART, for array associated reverse transcriptases. It includes a reverse transcriptase enzyme, a neighboring partner gene and a long array of repeating sequences. Anthropic says the layout is reminiscent of CRISPR arrays but stresses that the biological function of ART is still unknown.

How Claude arrived at the candidate

According to Anthropic, roughly 950 agents worked for about 21 hours using around 210 million tokens. The search covered about 1.9 billion protein clusters and more than 200,000 reverse transcriptases before narrowing thousands of candidates to a much smaller set for human review.

One agent noticed an unusual repeat pattern near a reverse transcriptase gene. Human researchers then tested the candidate in the laboratory and found that the repeat array is expressed as a collection of short RNA molecules.

Why this is not another CRISPR, at least not yet

The similarity to CRISPR is structural at this stage. There is no evidence yet that ART performs gene editing, provides antiviral immunity or can serve as a medical tool. Anthropic says experiments to determine the system's primary function are continuing.

That distinction matters because describing a system as CRISPR like can sound like a proven biotechnology breakthrough. At present this is an early biological finding that needs further study, scientific scrutiny and independent replication before its importance can be established.

What the work suggests about AI in science

The notable part of the project is the workflow: a general AI model was used to scan a vast search space, generate hypotheses and prioritize candidates, while humans carried out experimental validation. That suggests a way AI may expand scientific search without replacing laboratory work.

The announcement also comes from the company that develops Claude and describes early work. A fuller assessment will depend on a complete scientific publication, peer review, independent confirmation and a clearer explanation of what ART actually does.

Back to Latest News