Hacker News Reader

Hacker News

Comment by 323 | Hacker News Reader

Full thread

323

The sequence of the smallpox virus is publicly available:

https://www.ncbi.nlm.nih.gov/nuccore/9627521

What sort of labs have the skills to reverse genetics it into a cowpox backbone?

Are we talking 2-3 labs in the world, or basically any decent lab? Could a PhD in virology/microbiology do it?

View on HN

Any decent lab at an R1 university, or any bright and well motivated individual.

It doesn't take that much. If you're good enough at DNA build from oligos, could probably synthesize it with less than a PhD. It is expensive though, and very unlikely that anyone would actually do it in a way that evades the surveillance systems.

How much harder is it than the process of building your own bacteria from a DNA sequence...? Would just tacking it into a bacterial DNA sequence and being lucky end up in replication of it?

I remember when I looked into it that about $5k of equipment scrounged from ebay and stuff being trashed from labs was probably enough for making your own bacteria.

It's a bit more complicated, like how you can't just take a Mac binary, put it on a Windows computer and expected it to run just because it uses the same x86 machine code.

If you had the sequence, and cowpox, you could get mammalian cells to make.

But I think the hard part is not just synthesizing the DNA, but instead assembling the various proteins into an actual virus.

Can you tell a layman more about these surveillance systems?

Commercial oligonucleotide synthesis companies will check sequences submitted to them against databases of 'dangerous' or regulated sequences. Flagged sequences won't be synthesized. This was designed to prevent someone from ordering parts of a dangererous virus sequence for assembly of the complete virus.

Are the machines themselves treated like nuclear proliferation/dual use tech?

And if they are the next logical question is 'are the machines that make the machines themselves treated like like nuclear proliferation/dual use tech?'

I've assumed as much.

I've also wondered if there is some sort ability to order innocuous sequence A from company a, innocuous sequence B from company b and then in the privacy of your own lab do the DNA synthesis equivalent of A xor B to get dangerous sequence C.

I think it would be hard to get all the required fragments without raising red flags. There aren't that many synthesis companies and it would take a lot of fragments. Assuming you could get the fragments, its getting easier to assemble (join) them quickly. A 40000 base pair T7 bacteriophage genome was recently assembled from 52 parts in a single reaction (one tube)!

Fragment issues aside, the kind of stuff I'm thinking about is somewhat like homomorphic encryption at a genetic/biochemistry level to side step regulations.

> companies will check sequences submitted to them against databases of 'dangerous' or regulated sequences.

Weird, that's kind of how computer antivirus software works.

Reply on news.ycombinator.com