Phi X 174

So these were existing, an existing family.

They trained it on 15,000 different sequenced variants, just a few mutations to quite a few mutations, of this one virus that infects bacteria, Phi X174.

And then they had the AI spit out 300 designs for different mutated versions of that virus, and they tested them all, and 16 of them worked, which is an incredible success rate.

Now, doesn't necessarily mean that you can apply this to a human virus.

And they worked with a very simple, well-understood template.

The Phi X 174 virus.

Right, which only has about 5,400 to 6,000 base pairs.

Professor Martha Clokey, a prominent voice in the field, pointed out in our sources that the current results are highly constrained because of this simplicity.

These AI systems were trained on large collections of genetic sequences and then used to generate entirely new versions of bacteriophages, viruses that attack bacteria.

Researchers focused on a well-studied phage called Phi-X174 that infects E. coli.

The AI generated thousands of potential genomes, and researchers selected hundreds for laboratory testing.

Sixteen of the synthetic phages proved to be viable and capable of infecting bacteria.

I found, I, I found that surprising, and they, they don't talk about it, and maybe, maybe it's not surprising, but, you know, they say the genomic start sequences of all the Microviridae sequences were highly diverse.

All the PhiX174-like-

Yeah

... genomes started with the same consensus nucleotides.

One is the insertion of a new gene J in, in Evo63, extended non-coding regions, the loss of a gene, elongations of genes, swapping of genes.

But the origin of replication had no mutations again across all 16 genomes, so you really cannot change that at all.

And it, and at the phylogenetic level, these phages are similar to PhiX174-like microviridae, so that is what they wanted it to be.

Um, they, they then, uh, wanted to compare it to natural isolates of PhiX174, so they collected a bunch of data sets of natural PhiX-like viruses.

At the beginning of August, something pretty incredible happened.

Scientists published a paper detailing how they created new viruses using artificial intelligence.

Here's a new addition to the list of things AI can do.

Create new viruses not found in nature.

It does sound like something from a sci-fi movie, but this is real.

A U.S. team have used artificial intelligence to design completely new viruses.

The good news, they

say, the 16 new viruses created by AI pose no threat to humans.