Anthropic Claude sees CRISPR-like array in bacteriophage DNA

[.Scott]

TL;DR

Anthropic: "Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages."

Yesterday, issued a press release announcing (in part):

Today, we’re sharing early results from one of our first research programs, in which Claude autonomously discovered a novel enzyme system that is associated with an array of DNA repeats, a pattern reminiscent of CRISPR. Although we don’t yet know its function, the system that Claude discovered has a set of characteristics that have only ever been found together in a handful of other systems, all of which are programmable and perform operations like cutting, copying, and pasting DNA. Beyond CRISPR, which has already transformed science and medicine, several other such systems are now in development as promising tools.

There is a more formal paper at this link, a 40-page white paper. There is also mention of a "pre-print" - which may be that white paper.

If you go to this X.com page, you will find a post by Anthropic CEO Dario Amodei.
The first paragraph of that post is:

Today we announced the Claude-led discovery of a molecular machine that we suspect could represent a new gene editing mechanism. Its precise function, biotechnological utility (if any), or level of significance is not yet clear, but at minimum it is work I would have been proud to do as a PhD student. The work was done mostly, though not entirely, by Claude: our life sciences team suggested a broad area of research, Claude read through the literature and a bunch of genome data and discovered something interesting, then Claude proposed experiments to verify the discovery and our team carried them out.

The remainder of that post describe some current research and Claude's potential in biology - an advertisement for Claude.

Following that post (scrolling down) is one by @AnthropicAI. Since I have not join X.com, I only get this lead-in:

Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages. Beside the enzyme’s gene sits a long array of repeating DNA—a structure that looks somewhat similar to CRISPR. We don’t yet understand what this system does, but only a handful of known ...

So far, there's been a lot of reaction to the comment "it is work I would have been proud to do as a PhD student".
I have never pursued a PhD and it's been decades since I have been in an academic setting. I don't know what the criteria is for PhD work.
But certainly, when attempting to decipher working software from an unknown machine language, the type of pattern described by Claude is exactly what you need to identify. This is certainly an important find and I expect that continued investigation will uncover a new basic DNA mechanism - one that might be shared by more than just bacteriophages.

 

[BillTre]

My Question:
I don't know much about the Claude involvement, but it does not really surprise me it could do something like this, in the kind of data provided. I am assuming that a person decided that Claude should look for something like this rather than just it being the result of a random search among all of biological data and pulling out this relationship rather than some other relationship (that might be equally interesting).
Is this what happened?
Does the article say?

Some Biological Background:
This study indicates that: LUCA probably got the system from viruses, and that viruses used it to fight with other viruses that might have invaded the same cell (competition between viruses for the cell's resources). All very interesting.

The proposed CRISPR-like functional abilities of the viral sequence(s) is very interesting. Some consider CRISPR to have been present in LUCA (the Last Universal Common Ancestor). LUCA is oldest cellular structure that can be traced back to using phylogenetic methods (which are now traditional). Its thought that a LUCA encoded CRISPR was used to repel potential parasites (like viruses) that would have tried to invade these first cells (as it is in some modern cells). Subsequently, some evolutionary lineages derived from LUCA have lost CRISPR, and may possibly have evolved something else.
These are potential avenues of inheritance, as opposed to independent de novo generation of the sequences.

 

[Astronuc]

TL;DR: Anthropic: "Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages."

criteria is for PhD work.

Usually, a PhD candidate does 'original' work, which contributes the field of interest in the Arts and Sciences, e.g., in physics, chemistry, history, economics, biology, medicine, . . . . It could be a new theory/understanding or a novel application, e.g., a new measurement technique.

Similar to a MS degree, one demonstrates familiarity and understanding of the literature, but then one develops a new, i.e., original idea or approach, and often one is expected to publish several articles in relevant journals.

In the OP, the statement "Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages." indicates something new, because it was 'previously unknown'. Determining the enzymes function, biotechnological utility (if any), or level of significance, in addition to the discovery would be the basis of a PhD.

 

[.Scott]

In the OP, the statement "Claude has discovered a previously unknown enzyme system hidden in the DNA of bacteriophages." indicates something new, because it was 'previously unknown'. Determining the enzymes function, biotechnological utility (if any), or level of significance, in addition to the discovery would be the basis of a PhD.

Thanks for that info. As I said, it was topic number one for all of the initial readers of that press release. I presumed it to be an accurate criticism - but I take your post as authoritative.
My first reaction (shared to no one) to those initial posts that mentioned the PhD issue but nothing else was "spoken like true gate-keepers".

 

[BillTre]

There is lot of legitimate research that would not be the basis of a PhD. It also depends on who you are discussing it with.