I'm unfortunately going to give an equally hand-wavy response. There's nothing in there that strikes me as factually incorrect (I could be wrong though because I am not an expert on these viruses). I did a brief genome analysis myself - only ~5% of the arginines are encoded by the CGG codon. It's also true that those two introduce the furin site, and it's hypothesized that the furin site is critical to the viruses ability to enter the human cell. It's also true that it's a large difference between the closest variant and SARS-CoV 2. There's also the fact that it's an insertion, not a mutation. i.e. the gene acquired extra amino acids, not just changing the new ones (that's one point I think is hand-waved over in the OP's comment, as he points out that there are many mutations between SARS-2 and the closest relative - which is true, but there is (to my knowledge and I'm happy to be corrected on this) only one insertion, and that is the furin site bearing two adjecent non-canonical arginine codons next to each other that also introduce a restriction site.) However, I do believe the insertion is out of frame, which is odd. If I was engineering the protein, I'd probably make the insertion in-frame to reduce variables as much as possible. So that's a point against. The paper also just glosses over that.

The restriction site is interesting because to my knowledge, mammals don't produce the protein that would normally digest it (which implies that it's probably rare among infectious eukaryotic viruses), but again, I could be wrong there and am happy to be corrected. Typically, a restriction enzyme will, under the right conditions, cut the DNA (or RNA) at the restriction site. One of the interesting things here is that if I was introducing a restriction site to track GoF research, adding it directly in the thing I added greatly simplifies my life. If that restriction site goes away, I know I lost my insert. It's also nicer to use a restriction site because I can do the digest in 30 minutes on a benchtop, run an agarose gel in an hour, and know if I still have it after passaging the virus, vs say, sequencing, which is usually more expensive and takes longer. Especially if it's BSL2+ because now I need to put it over a BSL2+ sequencer.

It's a hypothesis. We'll never know. There's no conclusive evidence either way, and it's absolutely something we should all be talking about, and the scientists among us should be trying to properly falsify it to the best of our ability.

Thanks for that, I'm very glad I asked. Your hypothesis around the restriction site is rather eye opening. The scientists involved in this study have faced intense online campaigns against them.

https://twitter.com/Rossana38510044/status/13806444823669841...

> the scientists among us should be trying to properly falsify it to the best of our ability.

This is precisely why I am bothered this has been placed into "conspiracy theory" land, due to a lot of political reasons and a whole lot of online pressure with those "hand-wavy" arguments. The same type of arguments you originally did not accept. I get them constantly when I post facts, and the are meant specifically to try to steer the discussion away from anything I was pointing out, to get that last word in.

Did SARS-CoV-2, with its affinity for human ACE2, leak from a lab in Wuhan that was doing known gain of function experiments dealing with CoV and human ACE2?

Or did it somehow show up in a market right next to the lab, with "hand-wavy" explanations as to how that happened?

The science seems to strongly point to one of those. Since access to the lab has been completely shut down, and historical information on the web has been systematically being removed, finding a smoking gun is difficult. But I feel an extremely strong case can be made with circumstantial evidence.