Hacker News Reader

Comment by timr | Hacker News Reader

> Given the early epidemiological nexus of the wet markets there is every reason to believe this arose out of a reservoir in the fur or meat industry...

I can make the same argument for lab origin: "given the early epidemiological nexus of a government virology laboratory in the same city as the index cases, many of whom worked at the lab..."

The rest of your comment boils down to "science is hard and often ambiguous"...which it is, but is not an argument either way, and certainly not a reason to leap to one conclusion or the other.

BTW: the close relatives in Yunnan and Laos aren't actually that close. But it all depends on how you define "close", which just begs the original question. I wouldn't call the known relatives dispositive.

View on HN

An epidemiological nexus is much smaller than an entire city and the details of the early cases matter. Aside from the contact-tracing, investigators were able to get positive PCR recovery of sars-cov-2 from -environmental samples- at the wet markets, indicating an incredible degree of contamination there. If the nexus were in fact WIV we'd expect a very different pattern of early cases (eg conference travel - a lot of researchers from WIV were at a Nipah conference in Singapore just before the outbreak!)

You could of course argue that the authorities just covered up any damning links, but the early outbreak was pretty chaotic and information wasn't well controlled... and in any case at that point we're not longer capable of arguing based on any evidence at all.

My point is simply that a lack of fully verified lineage analysis is completely unsurprising and not at all evidence in favor of conspiracy.

I used to engineer viruses... By natural standards the Laos samples are extremely close - the mosaicity of coronaviruses in these bats complicate matters - but likely within a decade of a common ancestor to sars-cov-2. Closer than any other sample that we know WIV were working with beforehand (again, people claim they were hiding what they were doing, except there was no reason whatsoever to do so before the pandemic).

Though "not close" for sure in the sense that even if they did have these strains, you would never evolve sars-cov-2 from them in a dish, there simply wouldn't be enough passages to do so. And that's sort of the point - the evolutionary "risk surface area" of zoonosis is many orders of magnitude larger than this nutty lab-leak idea, and we know that these coronaviruses jump all the time from close animal contacts, having caused two other epidemics in the last 20 years...

> Aside from the contact-tracing, investigators were able to get positive PCR recovery of sars-cov-2 from -environmental samples- at the wet markets, indicating an incredible degree of contamination there.

So? It's a market. People go there, in large numbers. Is it really surprising that you'd find traces of a widely circulating respiratory virus in a place where large numbers of people congregate?

> By natural standards the Laos samples are extremely close

RaTG13 is closer than the BANAL viruses, and we already knew about it. You have to look at specific genes and squint and make hand-wavy arguments about how SARS-CoV2 evolved from recombination of ~5 different viruses to make any sort of claim that this is strong evidence of natural evolution.

I'll give you this: the RBD of BANAL-20-52 is the closest match yet for the RBD of SARS-CoV2. Still no furin cleavage site, though.

https://assets.researchsquare.com/files/rs-871965/v1/986c09c...

> the evolutionary "risk surface area" of zoonosis is many orders of magnitude larger than this nutty lab-leak idea

I love this. The evolutionary story is too complicated and improbable to imagine, but sure, the idea of making a few relatively simple additions to a known virus is "nutty". And hey...it's not like there was a lab in Wuhan writing grants about similar transformations or anything.

You have to be kidding. Look, I don't have a strong opinion one way or the other, but you're simply in love with a theory and dismissing all evidence to the contrary.

The furin cleavage site looks suspiciously like a hallmark cut-and-paste operation that is exactly what I would do as a genetic engineer.

Really? I'd like to hear more about this. What exactly looks so "suspicious" that it can't be explained by typical recombination events that are known to occur naturally in RNA viral genomes?

The beta coronavirus class has a paucity of naturally occurring furin cleavage sites relative to the other coronavirus clades. I'm just saying that copying sequence from genomes a few species over is very common strategy for engineering projects.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7836551/

Irony is that this paper is highly doctored to make it look (from the figures) like the opposite point, without outright lying, like they loaded the dendrogram of beta coronaviruses with a ton of sequences that are distinct but "basically from the same collection", to make the pie slice of betas with furin cleavage sites look much much bigger than it should be. Also, the betas are excerpted from the dendrogram of all coronaviruses...

Do you believe the DRASTIC leaked document?

https://drasticresearch.files.wordpress.com/2021/09/main-doc...

It's pretty complete, and exactly like what I would expect a DARPA/NIH grant to look like, having helped prepare one in my career (and one DOE grant). I can't imagine a group of nonexperts drafting this document - correctly - just to make up a conspiracy theory - I'm not a virologist but I am a molecular biologist and biochemist and skimming the leaked document all of the technobabble looks correct to my understanding of science.

umm. a paucity. So like, you admit that betacoronaviruses naturally do indeed sometimes have furin cleavage sites?

I have no opinion on DRASTIC.

> you admit that betacoronaviruses naturally do indeed sometimes have furin cleavage sites?

Never claimed otherwise. They are rare. The DRASTIC leak talks specifically about engineering cleavage sites into bat coronaviruses. (P. 11)

The trouble with the synthetic furin hypothesis and all this DRASTIC stuff on that grant, etc. is that all of this work would certainly be done in known viral vectors... you would never go messing around testing spike proteins in some totally random virus backbone whose in-vitro behavior hadn't been worked out before in your assay cell lines.

So, maybe I missed something, but none of the experimental designs I've seen in those grants would ever plausibly lead to the SARS-COV-2 sequence. Again, pre-pandemic there was no reason to be sneaky about this stuff if it were actually being done... I don't know why anyone would waste their time on some uncharacterized, completely markerless viral vector for these RBD and proteolysis tests.

Sigh... maybe if we -had- been doing those experiments earlier we would have discovered this new clade and been more ready for it. That grant looks spooky to some people precisely because it was so prophetic.

I don't think those grants would have led to the SARS-CoV-2 sequence either, because the base strain is different. BUT let's be real here. You don't ask for money and do the experiment. You do the experiment and then ask for continuation money. So it is a "generally accepted grantsmanship strategy" that WIV would have been working on viruses adjacent to that in the grant with the same techniques, and adjacent techniques that would lead to the natural next step. If that natural next step infected a postdoc that was careless that one time when he/she was pulling a 9-9-6 workweek... It's all so terrifyingly plausible to anyone who has been there on the front lines doing lab work.

You're really into stacking these hypotheticals. Why would anyone bother developing a pile of fiddly new vectors? cui bono? Vector development sucks, I've done it in simpler virus families. And why do it completely markerless? Remember that the other odd thing about this grant was that a lot of the genetic work was supposed to be done in the US. If it were already being done we'd probably have had those sequences in -our- databases!

My (recent) experience with viral development is that it can be done in radical multiplex. You don't make one vector at a time. You mix five together, let or encourage them to recombine, then use sequencing, screens and molecular assays to sort out the details of which of the trillions you create has your desired phenotype and why. If this group was thinking about adding a furin cleavage site, they could do so using a template homologous to the appropriate regions of diverse natural viruses, then use selection in cell culture or animal models to establish where it landed and what effects it had on infection.

It's highly unlikely that an established group would write a proposal that didn't describe their current research trajectory.

Although you have made a lot of arguments against it, I still don't know why it seems so unreasonable to you that SARS-CoV-2 might have laboratory origins. I think it's just that your prior expectations run against this. To me it's much easier to imagine than a natural spillover, given the modern urban circumstances and very unusual viral features and phenotype. Maybe my work experience is more aligned with this possibility than yours?

I specialized in high throughput synthetic biology and in applications adapting vectors and pathogenic proteins to subvert the human immune system (for immunooncology mostly). So I have all the background needed for supervillain thinking here :p Of course you can do multiplex markerless construction - but it’s a huge pain in the ass at these lengths and it still leaves open the question of what the point of such work would be.

I feel like I’m taking crazy beans - 20 years ago a satbecoronavirus from a wet market nearly started a global pandemic - hell my colleagues at the time were the ones that identified it as a coronavirus! We have epidemiological evidence this happened again. Absolutely nothing about this sequence is obviously synthetic or even unusual for coronaviruses. How is zoonosis not the null hypothesis?

> You do the experiment and then ask for continuation money.

Public research labs aren't made of money, and generally don't run ambitious experiments and then pray for funding to drop from the sky on them.

No, you use money/resources from a previous grant to do exploratory work on the next grant while you are writing it. Sometimes you double-dip, so you write up a grant for one place, change parameters and personell slightly for another, etc. Have you ever actually worked in a research lab?

Nobody seriously believes that COVID-19 could have been produced via exploratory work.

I wouldn't believe it if you told me COVID-19 was produced via carefully planned work. IMO, COVID-19 could only have been produced via exploratory work.

Look, fact of the matter is, I have consumed ~2M USD of resources on a seven year PhD that did not have grant money attached to it. Despite my important results, my grad school PI did not pursue further work in the direction I piloted. And as a postdoc I've steered DOE funded bioengineering project away from a strategy I knew would be unfruitful towards results that garnered three papers in as many years, but were very much not "in the proposed work of the grant" (though very much in the spirit of it). Ultimately we rewrote the grant to reflect the effect magnification I got and we were rejected for renewal, haha. Moreover, at the end, we did the experiment we were actually funded for, and found no effect.

Grants and science writ large are not the ideal spherical cow you think it to be.

Per the CCDC 33 of 585 environmental samples taken early Jan. were positive, and 31 of those were concentrated near the live-animal region of the Huanan market. Environmental PCRs aren't magic, they have a finite level-of-detection, RNA viruses don't survive very long in the wild...

There's no handwaving and I'm not positing a particular evolutionary history - the sarbecoronaviruses are clearly understudied and we don't have a good survey of what's out there, we just know about a few close relatives. These are known to be highly mosaic viruses!

"few relatively simple additions to a known virus" Which additions and which known virus? There are a huge number of random mutations between the above mentioned strains and SARS-COV-2, there's no plausible story for lab manipulation here from those strains. ...have you ever worked with viruses before?

There are billions of animals in the chinese markets, each a test tube... this is the cauldron of evolution, not a few dinky petri dishes in one lab, man.

I'm in love with parsimony, and the history of human disease is the history of zoonosis. If extraordinary claims are going to be made, the burden is on those with the radical new theory to offer up a shred of molecular or epidemiological evidence, not just stories about conspiracies.

Recent work re-evaluating timelines has only -strengthened- the case for Huanan market and a possible link to a racoon-dog intermediate. https://www.science.org/doi/10.1126/science.abm4454

The trouble with that piece, is that it is contradicted by this:

https://www.reddit.com/r/China_Flu/comments/es21aq/wuhan_pne...

(Chinese original has been deleted, but I remember seeing it at the time and running it through google, scroll down for English translation in the comments)

which reports a case with symptoms starting 24th December, and reports of the hospitals in Wuhan already having multiple cases. Patient didn't visit the market, but did work near.

With what we now know of disease progression, and infectiousness, this pushes the initial date back at least a month, probably two or more.

...but that report doesn't contradict anything. The current earliest known case is a Huanan market seafood vendor who contracted the disease on Dec -11-. Dec 24 is way late in the existing epidemiological timeline. (And no one is saying the above is the provable first case, just the first recorded case.)

No, you're wrong. The earliest cases were at least in November, and quite probably earlier.

The earliest case identified in the market was on December 11. There's simply no way that was the first actual case.

https://twitter.com/BallouxFrancois/status/14621732432474071...

If you want to argue like that there's also this publication from Italy claiming that the first antibodies were detected in September 2019.

The comment above says I would just do X, and bringing up RaTG13 and BANAL-52 as an argument as if it's all trivial. They share 96.2% and 96.9% similarity with sars-cov-2. There are seemingly unrelated animals sharing that much. The linked study also doesn't mention any of that.

I'm not going to argue one way or the other, but I'm starting to get really annoyed at people posting random studies they half understand to argue one way or the other. Covid-19 discussions hackernews(or reddit for that matter) are like the Joe Rogan of nerds. It's ridiculous, people here should know or act better.

https://pubmed.ncbi.nlm.nih.gov/33176598/

One strain of BANAL (52) has a high DNA sequence similarity to SARS-CoV2. Nonetheless, RaTG13 is closer overall (as you can see in the phylogenetic analysis in the paper I posted), and there are enough differences that you have to make complex assumptions about how multiple viruses crossed over evolutionary time. Possible? Yes. Parsimonious? Not necessarily.

Also, nobody reasonable is arguing that BANAL-52 is not evolutionarily related to SARS-CoV2. Of course it is. But that could be true even if it were a precursor of some virus that was later engineered in a lab.

To fully convince me of a natural origin hypothesis, you'd have to find a virus in an animal reservoir that had a highly similar RBD with a furin cleavage site -- or something within a few point mutations of a furin cleavage site. Also, a high degree of homology across the remainder of the genome.

I'm fine with people posting random studies that I can read and form my own opinion on. This isn't ridiculous. This is the process of science.

I’m actually not fine with people posting random studies and then also making random claims from undereducated interpretation. That’s not the process of science, mostly because in the process of science you’re expected to be educated first before you can enter a discussion, otherwise you’re not discussing heavily niche subjects with peers you’re just introducing noise and that drowns out the actually useful educated niche. As a layman that is trying to find what the actual, educated in virus research and virus forensics folks are discussing and saying, this sort of undereducated claim is more noise for me to try and sift. It’s very annoying and a waste of my time.

"Chinese original has been deleted"

Wait that never happens...

> Per the CCDC 33 of 585 environmental samples taken early Jan. were positive,

The first cases were at least in November, and quite probably earlier. The virus was in circulation in Northern Italy by December. We know this from multiple lines of evidence.

https://twitter.com/BallouxFrancois/status/14621732432474071...

Are you seriously posting a link to a one-off phylogenetic molecular clock estimate as experimental evidence of italian priority? The other Italian claims of priority are highly disputed by basically everyone - but have I missed anything - are there any actual sequenced, dated samples from Italy in December? That's the standard of evidence here, not some crappy antigen test or unsequenced amplification hit. No one believes the Dec 11 case to be "the first", just the first well attested case.

No. I'm seriously posting a paper by the director of the University College of London Genetics Institute.

The same twitter thread (again, same person: director of UCL UGI) shows a Lancet article placing the first documented hospitalization on December 1, and a different source documenting a case in China in mid-November:

https://www.scmp.com/news/china/society/article/3074991/coro...

So yes, there are multiple lines of evidence. This isn't even remotely controversial. You're arguing that the sky is red, and your only counter-argument to the evidence otherwise is incredulity.

> are there any actual sequenced, dated samples from Italy in December? That's the standard of evidence here, not some crappy antigen test or unsequenced amplification hit.

I mean, you're inventing "standards of evidence" here, but as long as you're asking: yes. Dated, PCR-confirmed wastewater samples in Italy were purified by gel electrophoresis and sequenced.

https://twitter.com/BallouxFrancois/status/13280584020529684...

https://www.medrxiv.org/content/10.1101/2020.06.25.20140061v...

I'm not a huge fan of "appeal to authority", but I'm curious what your credentials and experience in this area are?

The person you are arguing with claims to have direct experience working with viruses in a lab setting.

Is this another of HN's all too frequent, "computer science major arguing with expert in field completely unrelated to computer science"?

Reasoning doesn't require credentials, please let's keep it that way.

One of the parties did an appeal to authority as a way to _prove_ they entered _relevant_ evidence.

This is obviously entirely misguided, and if you start making such basic logical mistakes it is expedient to inquire about the posters background to know if they've reached the limit to their ability to create reasoned arguments.

There are nuances, but at it's core either it was valid for them to make that appeal and it is valid for us to request their credentials (which is essentially an appeal to authority--"who has the better credentials?"), or it was not valid for them to make that appeal and it is not valid for us to request their credentials. I lean towards the latter view.

Logical mistakes don't weaken a person's entire argument (we don't want to just see who's "winning"), or even weaken the line of reasoning they were invoked to support -- they just fail to provide evidence.

One of the parties called into question the validity of a paper the other party posted, and the other party rebutted by implying that as the director of an established institution, the paper's author has something to lose. I find that argument weak, but the initial questioning of validity also provided no evidence or made any strong arguments as to why the paper was not valid.

Appeals to authority are useful not as logical arguments but as a way to avoid wasting time on drivel, and I don't believe that to be the case here.

No one is arguing that the earliest known case on Dec 11 is when sars-cov-2 "began". We know it must be older than that. More recent reviews of chinese epidemiology don't mention the hubei patient, the Dec 11 case seems to be consensus earliest verified case.

And yes, for a data point of this importance a random environmental sewage PCR handled over half a year later is of dubious provenance. If you've never done a ton of environmental PCR work you'd be amazed how easy it is to contaminate everything, but whatever. Given how contagious sars-cov-2 is, if it really were present in Italy in early December... we'd probably have more definitive proof of it's presence much earlier than Jan 31.

Ok, fine. We're agreed then: the virus was obviously circulating widely by December, which means in all likelihood, it could have been found in pubic places in Wuhan China. Like, say...a public market.

I'm not saying it had to happen that way, just that finding shreds of viral RNA in a market, in December, in Wuhan, is in no way dispositive evidence of the market being the origin of the virus -- any more than finding flu virus in the ball pit at a McDonalds means that the flu came from Ronald McDonald.

> And yes, for a data point of this importance a random environmental sewage PCR handled over half a year later is of dubious provenance. If you've never done a ton of environmental PCR work you'd be amazed how easy it is to contaminate everything, but whatever.

"They could have made a mistake, so I'm going to ignore all evidence which disagrees with my prior beliefs."

Millions of old people in dense cities turn out to be a fantastically sensitive instrument for detecting the presence of sars-cov-2 in a population. You’d have to explain nearly two months of cryptic transmission that left no hospital reports or verifiable samples…

You do realize these papers pointing to an international nov-dec timeline only -weaken- any link to WIV at all, and have in fact been heavily leaned upon by Chinese conspiracy theorists as evidence this disease has -western- origins?

> You do realize these papers pointing to an international nov-dec timeline only -weaken- any link to WIV at all, and have in fact been heavily leaned upon by Chinese conspiracy theorists as evidence this disease has -western- origins?

Chairman Pooh? Is that you?

The CCP can fuck right off with this idiocy. Seriously. Nobody with half a brain is falling for it.

You don't get to suppress information and/or not look for it in the first place, then claim that the lack of information is indicative of something.

Then why did the Chinese shut down access to the covid database for instance if there is nothing there? Edit : You can downvote this, but that doesn’t answer this question that I am seriously curious about?

The same reason they surpressed all evidence of widespread wildlife trade in the city in 2019 and lied to the WHO about it?

> If extraordinary claims are going to be made ...

I don't think any claim here is extraordinary, and none with very good evidence.

So pursuit of parsimony doesn't help much.

Thank you for bringing some perspective to the discussion, at the same time, I have no illusion that it is going to help to convince people that want to believe in the lab leak theory of anything else. Wanting to assign blame for all this is a very strong desire in many and the Chinese government through its acts and omissions has unfortunately made it a lot easier for people to continue to be deluded about this. Some of the arguments made are so incredibly weak that they shouldn't be made at all, and yet, they are supporting this whole house of cards of criminal intent gone wrong. It's a pretty sad statement as to the degree of understanding that people in general have about these matters, if they studied the subject (by reading textbooks, not watching youtube videos) we might be able to sidestep some of this.

> Wanting to assign blame for all this is a very strong desire in many and the Chinese government through its acts and omissions has unfortunately made it a lot easier for people to continue to be deluded about this. Some of the arguments made are so incredibly weak that they shouldn't be made at all, and yet, they are supporting this whole house of cards of criminal intent gone wrong.

Nobody so far in the thread has said anything about "criminal intent". That's just you, interjecting your pre-existing biases into the discussion.

It's entirely possible that a laboratory was doing completely legal experiments with no malicious intent, and an accident happened. In fact, it's so possible, that reasonable people who understand the evidence have concluded that we're unlikely to ever know the truth.

That's where I sit on the matter.

> It's a pretty sad statement as to the degree of understanding that people in general have about these matters, if they studied the subject (by reading textbooks, not watching youtube videos) we might be able to sidestep some of this.

Maybe you should look at my profile.

Thanks - I don't expect to change anyone's mind. I just don't want to give into the cynicism that evidential reasoning is pointless. I don't want to tell anyone what to think... but given that I have the professional background, I feel some moral duty to at least try to outline why most biologists assign low probability to a lab-leak.

I think one big missing element for the reasoning behind all this is that people seem to intuitively assign the wrong prior to the chances of a zoonotic event, they miss that fully 2/3rds or so of our pathogens are zoonotic in origin, and that's only the ones that we can trace back, the remainder is either an open question or settled to have originated in our species or passed on as part of our heritage prior to humanity budding off from our ancestors.

I'm uncertain why it does not seem to be discussed that maybe the leak was sabotage by a lab worker who got paid really well by a foreign service to take it from the lab and disperse it.

Environmental samples from the wet market only show that infected humans were there. Given the centrality of markets in human social networks it seems very normal that an outbreak in the city would eventually focalize in the market. Without samples from animals in the market, the evidence for spillover happening there is extremely weak, and not even circumstantial, because it's indistinguishable from what we would expect given a spillover anywhere in Wuhan.

The problem with relying so heavily on argument from coincidences is that one then incurs the epistemic responsibility to seek out and explain alway other coincidences in the neighborhood putatively pointing to alternative explanations (a step people often forget, but is essential for the validity (insofar as one can speak of validity) of one's argument from coincidences). And such coincidences are always easy to find, since coincidences are cheap. (To take just one example, the pandemic happened shortly after the World Military Games, which US servicemen attended, at a time when the relations between the two countries were at their worst in recent times, and who came from a country one of whose household names novelists once publicly fantasized about all the world's nations coming together and use biological weapons to wipe out the entire population of China, based on none other than the desire to curb China's peaceful rise (cf. Unparalleled Invasion)) Do these "coincidences" need explaining away? Do they need to create a presumption in favor of the bioattack theory? If not, what is it about spatial proximity coincidence that gives it such power to confer that sort of presumption?

including the strain from the outbreak that the batwoman published in 2015 or so?

That one was supposedly the closest.

Reply on news.ycombinator.com