Is SARS-CoV-2 / COVID-19 a lab escapee?
October 21, 2022 5:56 AM Subscribe
A paper entitled Endonuclease fingerprint indicates a synthetic origin of SARS-CoV-2 by Washburne et.al. examines the structure of the COVID-19 virus strains and finds a strong suggestion that it was snapped together in a lab like LEGO. Lay explanation, and a Twitter thread summary. This is a pre-print; the authors claim rigour and invite critique and reproduction from others. Nothing nefarious is inferred.
This post was deleted for the following reason: Preprint. -- loup
I question the objectivity of the paper. Beyond the rather techno-intense language of the "lay" explanation,
The technology used to make infectious clones is relatively cheap, especially compared to making an atom bomb. Even if our theory is rejected by later tests, the ease of these experiments should scare the shit out of all of us enough to start talking about global biosafety.... I want to reiterate: we don’t know ‘whodunit’. I have been very vocal on the topic of COVID origins, and it’s essential to separate this research paper from my public proclamations.
Very strong "I know I've been yelling about this for a long time, but really - this time I'm serious" vibes.
posted by Silvery Fish at 6:08 AM on October 21, 2022 [4 favorites]
The technology used to make infectious clones is relatively cheap, especially compared to making an atom bomb. Even if our theory is rejected by later tests, the ease of these experiments should scare the shit out of all of us enough to start talking about global biosafety.... I want to reiterate: we don’t know ‘whodunit’. I have been very vocal on the topic of COVID origins, and it’s essential to separate this research paper from my public proclamations.
Very strong "I know I've been yelling about this for a long time, but really - this time I'm serious" vibes.
posted by Silvery Fish at 6:08 AM on October 21, 2022 [4 favorites]
I'm as critical of peer review and the biases of academic publishing as anyone else who's spent time in academia, but taking a preprint that's this incendinary, writing up a pop-science article on it and promoting it on social media is irresponsible at best. The decision to do that makes it very clear that the priorities of the people behind it are not with making sure the conclusion is sound.
posted by automatronic at 6:09 AM on October 21, 2022 [18 favorites]
posted by automatronic at 6:09 AM on October 21, 2022 [18 favorites]
The technology used to make infectious clones is relatively cheap, especially compared to making an atom bomb.
This comparison seems like an utter non-sequitur intended to suggest an idea without saying it - "this is a weapon". Not great science writing.
posted by Dysk at 6:13 AM on October 21, 2022 [6 favorites]
This comparison seems like an utter non-sequitur intended to suggest an idea without saying it - "this is a weapon". Not great science writing.
posted by Dysk at 6:13 AM on October 21, 2022 [6 favorites]
If it's good enough for Jon Stewart...
posted by They sucked his brains out! at 6:13 AM on October 21, 2022
posted by They sucked his brains out! at 6:13 AM on October 21, 2022
it’s essential to separate this research paper from my public proclamations.
We can't, though. That's not how this works.
posted by mhoye at 6:19 AM on October 21, 2022 [2 favorites]
We can't, though. That's not how this works.
posted by mhoye at 6:19 AM on October 21, 2022 [2 favorites]
This is, charitably, not responsible science and while I'm not in charge of anything here I feel like we should not be amplifying it. People who genuinely "claim rigour and invite critique" aren't doing that on Twitter.
posted by mhoye at 6:21 AM on October 21, 2022 [13 favorites]
posted by mhoye at 6:21 AM on October 21, 2022 [13 favorites]
Nothing nefarious is inferred.
The technology used to make infectious clones is relatively cheap, especially compared to making an atom bomb.
Um.
posted by Halloween Jack at 6:22 AM on October 21, 2022 [3 favorites]
The technology used to make infectious clones is relatively cheap, especially compared to making an atom bomb.
Um.
posted by Halloween Jack at 6:22 AM on October 21, 2022 [3 favorites]
It's been a while, but I'm pretty sure the whole point of Golden Gate is to remove the enzyme recognition sites during cleavage, so that the sequence can't come apart during later digests (as you link up additional pieces). Leaving them in like this seems rather odd, because you can only assemble the final product in one shot (otherwise the next digest will dice your sequence into half a dozen pieces).
As well, the BANAL-116 sequence (which came from a cave in Laos) has fewer BsaI sites than even the initial pandemic sequence, which undermines their point a fair bit. Fig 3B shows a bunch of disparate restriction site patterns, and it sure feels like they're cherrypicking to make their data feel stronger than it is.
posted by Orange Pamplemousse at 6:23 AM on October 21, 2022 [5 favorites]
As well, the BANAL-116 sequence (which came from a cave in Laos) has fewer BsaI sites than even the initial pandemic sequence, which undermines their point a fair bit. Fig 3B shows a bunch of disparate restriction site patterns, and it sure feels like they're cherrypicking to make their data feel stronger than it is.
posted by Orange Pamplemousse at 6:23 AM on October 21, 2022 [5 favorites]
I dunno, I just really, really doubt all the "lab escapee" stuff that has come out since the very start of the pandemic. None of it has born fruit so far, but my main source of doubt as a non-virologist is this: diseases usually require multiple introductions to really take off. Initially, covid was infectious but not nearly as infectious as it is now, and it's difficult for me to believe that one accidental lab introduction could have produced a global pandemic. We've had infections from lab error before and they have fizzled out.
That's why, again as a non-virologist, I find the animal reservoir idea more persuasive - it allows for multiple introductions over time until there are a lot of human cases locally and the disease can take off.
posted by Frowner at 6:27 AM on October 21, 2022 [6 favorites]
That's why, again as a non-virologist, I find the animal reservoir idea more persuasive - it allows for multiple introductions over time until there are a lot of human cases locally and the disease can take off.
posted by Frowner at 6:27 AM on October 21, 2022 [6 favorites]
So how is this not just the academic-washed version of shouting "kung flu"?
posted by AlSweigart at 6:27 AM on October 21, 2022 [3 favorites]
posted by AlSweigart at 6:27 AM on October 21, 2022 [3 favorites]
From a quick read of the linked twitter thread, the argument appears to be:
- We identified a property of viruses that can be quantified
- We estimated the distribution of that property's values among naturally occurring viruses
- sarscov2 has one of the less likely values.
- We conclude that sarscov2 is a candidate for an engineered virus.
- We then tested only sarscov2 for some other properties which we claim are hallmarks of engineered viruses, and the results reinforce that conclusion
It seems like a flaw that they ran confirmatory tests but only on sarscov2. The fact that it's an outlier on the first measure is not conclusive even a little bit. It is, after all, an outlier in many other ways too. So without doing those same confirmatory tests on other strains, the argument just doesn't seem very strong to me.
posted by dbx at 6:27 AM on October 21, 2022 [6 favorites]
- We identified a property of viruses that can be quantified
- We estimated the distribution of that property's values among naturally occurring viruses
- sarscov2 has one of the less likely values.
- We conclude that sarscov2 is a candidate for an engineered virus.
- We then tested only sarscov2 for some other properties which we claim are hallmarks of engineered viruses, and the results reinforce that conclusion
It seems like a flaw that they ran confirmatory tests but only on sarscov2. The fact that it's an outlier on the first measure is not conclusive even a little bit. It is, after all, an outlier in many other ways too. So without doing those same confirmatory tests on other strains, the argument just doesn't seem very strong to me.
posted by dbx at 6:27 AM on October 21, 2022 [6 favorites]
n.b. I posted it not as an endorsement but as a subject of conversation, and specifically to get concentrated commentary on it from people who actually know what they're talking about.
posted by seanmpuckett at 6:29 AM on October 21, 2022 [1 favorite]
posted by seanmpuckett at 6:29 AM on October 21, 2022 [1 favorite]
"Then a miracle happened....."
posted by pthomas745 at 6:31 AM on October 21, 2022 [1 favorite]
posted by pthomas745 at 6:31 AM on October 21, 2022 [1 favorite]
I posted it not as an endorsement but as a subject of conversation, and specifically to get concentrated commentary on it from people who actually know what they're talking about.
The problem is that this post is indistinguishable from COVID disinformation spread under the guise of "just asking questions".
So that is what I'm going to call it.
posted by AlSweigart at 6:38 AM on October 21, 2022 [14 favorites]
The problem is that this post is indistinguishable from COVID disinformation spread under the guise of "just asking questions".
So that is what I'm going to call it.
posted by AlSweigart at 6:38 AM on October 21, 2022 [14 favorites]
Widely circulating pre-prints on social media is like declaring election winners or football winners 30% before the end.
Another obvious issue is the extreme conflict of interest pushed upon all modern scientists: one of the quickest ways to launch a career or get a promotion is by publishing incendiary/flashy/heterodox papers. This is to some extent natural, as Popperian science progresses via continual undermining, in the pursuit of that which cannot be undermined.
But, we don't have to give this shit air space, any more than we need to hear "both sides" of the arguments about racism etc. What a lot of non-scientists don't get is that professionals in the relevant field won't really buy these kinds of edge-lord results until A) it has been fully published in a good journal B) a few years have passed and nobody has managed to knock it down, and C) Some sort of follow-up or corroboration work has been published that shores up the results of the shaky first paper.
posted by SaltySalticid at 6:45 AM on October 21, 2022 [2 favorites]
Another obvious issue is the extreme conflict of interest pushed upon all modern scientists: one of the quickest ways to launch a career or get a promotion is by publishing incendiary/flashy/heterodox papers. This is to some extent natural, as Popperian science progresses via continual undermining, in the pursuit of that which cannot be undermined.
But, we don't have to give this shit air space, any more than we need to hear "both sides" of the arguments about racism etc. What a lot of non-scientists don't get is that professionals in the relevant field won't really buy these kinds of edge-lord results until A) it has been fully published in a good journal B) a few years have passed and nobody has managed to knock it down, and C) Some sort of follow-up or corroboration work has been published that shores up the results of the shaky first paper.
posted by SaltySalticid at 6:45 AM on October 21, 2022 [2 favorites]
We've had infections from lab error before and they have fizzled out.
A large percentage of people infected with Covid-19 do not develop symptoms, while others die. Totally different from a virus where you're accidentally stuck with a needle and are essentially dead already.
posted by jabah at 6:51 AM on October 21, 2022
A large percentage of people infected with Covid-19 do not develop symptoms, while others die. Totally different from a virus where you're accidentally stuck with a needle and are essentially dead already.
posted by jabah at 6:51 AM on October 21, 2022
A failure of this preprint (meaning that it has not been formally reviewed by outside researchers) is that it ignores both virology and evolution. Luckily, your local friendly evolutionary microbiologist missed his train and gets to sit around for the next 30 minutes at the station with nothing to do.
There are two glaring issues after a quick scan of this. First, the authors compare the CoV2 genome with 70 other coronavirus genomes; they bioinformatically broke apart these 70 genomes to find out distributions of these genomes, and then compared this distribution with CoV2. This technique works well as long as the distribution comes from a random sample. But, these 70 genomes are not truly random — they are all related to each other through evolution (hence, why they are all called coronaviruses!). We need to correct for this evolution to make these characteristics a true distribution that we can compare. The authors did not.
The paper also looks at silent mutations — changes in the DNA (or RNA in this context) that do not change the protein sequence. There are 4^3 = 64 sets of three DNA nucleotides, but these 64 codons code for 21 possibilities, so they must be some redundancy moving from DNA to proteins. Biologists learn this in intro / genetics / molecular evolution class. Researchers who work on viruses know that this is baloney — any change in the DNA / RNA is going to affect dynamics inside of the cell, and so silent mutations suddenly have a huge impact. RNA binds to itself to create fun shapes that determine how well the molecule is going to function, so silent mutations are suddenly not so silent, and we’ll see this in how the genome is selected for through time. This effect is also pushed along by how many viruses there are in a host: evolution works differently on a set of 10,000 individuals (think of prehistoric humans) compared to a set of 100 million individuals (think of number of viruses in an infected individual — and then multiple this number by the number of infected humans in the planet).
This paper needs the expertise of an evolutionary biologist and/or a virologist. I found these two issues after a 4 minute scan. Someone with these areas of expertise needs to contribute to this paper with 4+ hours of thinking about it. The fact that all three coauthors do not have a background in these areas AND that they did not invite such a person to contribute to their paper is not great science.
posted by Peter Petridish at 6:51 AM on October 21, 2022 [14 favorites]
There are two glaring issues after a quick scan of this. First, the authors compare the CoV2 genome with 70 other coronavirus genomes; they bioinformatically broke apart these 70 genomes to find out distributions of these genomes, and then compared this distribution with CoV2. This technique works well as long as the distribution comes from a random sample. But, these 70 genomes are not truly random — they are all related to each other through evolution (hence, why they are all called coronaviruses!). We need to correct for this evolution to make these characteristics a true distribution that we can compare. The authors did not.
The paper also looks at silent mutations — changes in the DNA (or RNA in this context) that do not change the protein sequence. There are 4^3 = 64 sets of three DNA nucleotides, but these 64 codons code for 21 possibilities, so they must be some redundancy moving from DNA to proteins. Biologists learn this in intro / genetics / molecular evolution class. Researchers who work on viruses know that this is baloney — any change in the DNA / RNA is going to affect dynamics inside of the cell, and so silent mutations suddenly have a huge impact. RNA binds to itself to create fun shapes that determine how well the molecule is going to function, so silent mutations are suddenly not so silent, and we’ll see this in how the genome is selected for through time. This effect is also pushed along by how many viruses there are in a host: evolution works differently on a set of 10,000 individuals (think of prehistoric humans) compared to a set of 100 million individuals (think of number of viruses in an infected individual — and then multiple this number by the number of infected humans in the planet).
This paper needs the expertise of an evolutionary biologist and/or a virologist. I found these two issues after a 4 minute scan. Someone with these areas of expertise needs to contribute to this paper with 4+ hours of thinking about it. The fact that all three coauthors do not have a background in these areas AND that they did not invite such a person to contribute to their paper is not great science.
posted by Peter Petridish at 6:51 AM on October 21, 2022 [14 favorites]
« Older Guardian emus ferocious with locusts and foxes... | The Firebird is the Word Newer »
This thread has been archived and is closed to new comments
posted by 3.2.3 at 6:02 AM on October 21, 2022 [4 favorites]