Thread by @acritschristoph

Alright, I had a look last night and will quickly do a thread on it.
There are many kinds of 'wrong' in science, but this preprint is False. There are many reasons (links at end), but the main one: the “unusual” sites are all *exactly* found in natural bat coronaviruses. 1/n

https://twitter.com/BallouxFrancois/status/1583165259799412737

The authors focus on cut sites of BsaI and BsmIB. Why these two REs? The simple answer is there is no good reason- they were just cherry-picked as the “most unusual”. Here's a plot of just a few RE cut sites possibilities across coronavirus genomes (SARS-CoV-2 at top)

And here are the two they happened to choose. Not because they are the most commonly used - or the best choice - but post-hoc rationalized because one of them was used previously in an old coronavirus paper from UNC.

OK, so what does the preprint say? So does SARS-CoV-2 have more RE cut sites than natural CoVs? No. Does it have more BsaI and BsmIB cu5 sites? Also no! Are they in convenient spots for engineering, like the edges of genes or around the RBD or spike gene? Again no.

At this point you might ask what *is* unusual about them according to the preprint? Simply that they are spaced evenly so no fragment would be more than 8 Kb- that's the measure of “synthetic-ness”. Not exactly hard with 5 sites in 30 Kb, after cherry picking an enzyme.

They develop a score for how evenly spaced the fragments are. Hilariously, their analysis *literally shows* that other natural CoVs score higher than SARS-CoV-2 on this metric and thus look “more engineered”. Clearly it's not exactly a good signal of artificial manipulation!

OK- so it's (a) cherry picking of (b) a signal that doesn't scream 'engineering' that (c) isn't even more extreme than some other natural viruses.
And now this is the point where we note every single one of the sites [or lack of] also occurs in a natural coronavirus as well.

So the authors must be proposing that someone added and removed RE sites to exactly match sequences of natural coronaviruses that were not even discovered until 2020-2022. Why? No idea, but also the theoretical experiment being proposed never made any sense to begin with, so 🤷‍♂️.

So what actually happened? SARS-CoV-2 shares some of this 'pattern' with some viruses, but not the whole thing with any one other virus. That's true generally of its genome. Why? We all know the answer: recombination.

@jepekar previously inferred the recombinant ancestral sequence of SARS-CoV- called 'recCA'. And we can check for the pattern of restriction sites in this sequence. Result: it shares almost the exact pattern of cut sites that SARS-CoV-2 has, minus a single mutation.

In the figure above, you can also see other bat viruses do have the same mutation as SARS-CoV-2 at this final site, clearly indicating it is a site that naturally changes frequently.

The RecCA result should follow intuitively from the fact that other viruses have these sites: recombination is why SARS-COV-2 shares these sites with other viruses in the first place. But phylogenetics clearly shows they were naturally inherited from these viruses.

What about missing sites? The authors propose that someone made a bizarre combination of additions and deletions of cut sites. RecCA matches SARS-CoV-2 at all missing sites because other viruses do. E.g. this one: similar to RpYN06 not just at the mutation, but the entire region.

So clearly that is a recombinant event, and not an engineering event. And the phylogenetic modeling behind RecCA confirms it. This is all as close to genomic "proof" that the preprint is flat out wrong as you can get when it comes to "proof" in science.

Many others have pointed out many, many issues with this work: the above demonstrates it is factually incorrect, but it doesn't make sense at any level. Amplifying a few:

https://twitter.com/K_G_Andersen/status/1583252858740219906

https://twitter.com/stgoldst/status/1583230301475438592

https://twitter.com/alchemytoday/status/1583213235737878531

https://twitter.com/wanderer_jasnah/status/1583285228587864064

https://twitter.com/zhihuachen/status/1583383858455916546

https://twitter.com/flodebarre/status/1583424593762537475

https://twitter.com/alchemytoday/status/1583361758903013376

https://twitter.com/TheEthanIverson/status/1583188380279803904

https://twitter.com/TheEthanIverson/status/1583430533727457281

Reflection time: how did we get here? Most experts find this preprint so embarrassingly bad that they refuse to comment on it; others don't have the time to look into the details, and say maybe; and yet some scientists comment positively out of ignorance (@BallouxFrancois)

As scientists we participate in communication platforms that fail so utterly that millions have already seen this obviously flawed work and will never hear it's wrong. More people have seen this straightforward lie and internalized it as true than maybe any paper you'll write.

We saw this a long time ago with the HIV inserts preprint. This time it seems like the system worked even worse! Post-COVID there is a cottage industry of scientists suckered into conspiracy mumbo-jumbo and egged on by bots and their retweets. Is there much hope for scicomm? IDK

I see the account above was (I presume temporarily) deleted - originally this tweet referred to the tweet in this screenshot. You can also access the original preprint here if the personal desire emerges

https://twitter.com/WashburneAlex/status/1583145276151189504

Appending additional threads of value here from the molecular perspective:

https://twitter.com/santiagokique/status/1583400895160590337

https://twitter.com/Friedemann1/status/1583520013864308736