Lab leak - pro or con

Here I have compiled a point-by-point list of the arguments for and against the lab leak hypothesis(es).

Sources and explanations can be found in the full blog post, but this may be easier to read if you don't feel like reading any more.

I also address Sigrid Bratlie's own hypothesis about what happened and review all ten of her points.

Arguments against lab leak and for zoonosis

What are the arguments against the lab leak hypothesis?

1. There is no evidence that there were ever sick employees at the Wuhan lab.
2. There is strong evidence that there were two separate animal-to-human transmissions, the so-called A and B lineages. This is not compatible with a lab leak.
3. The spread pattern of early infections based on many different sources and analyses all show a cluster around the Huanan wet market in Wuhan.
4. Environmental samples from the wet market show that the SARS2 virus was present there, with exactly the genetic variation expected from the earliest variants. Most positive samples were also from the part of the market where the relevant animal species were kept captive.
5. Blood bank analyses of 40,000 blood samples from Chinese hospitals collected in autumn 2019 and early 2020 in Wuhan found no infected individuals before December 2019.
   
   This is a classic example of conspiratorial reasoning, because these samples were demanded by lab leak supporters who claimed that the Chinese had to release them in the hope that they would show infection earlier in the autumn, which could coincide with their claims about lockdowns, infected soldiers, infected employees, etc. from September and October 2019. But when the data was released and showed that there was no infection before December, it was no longer mentioned.
   
   It doesn't fit their preferred hypothesis, so we keep quiet about it. The same goes for Bratlie, who never mentions this in her book. But then again, she only trusts Chinese data when it supports her hypothesis...
6. The super-spreader hypothesis also assumes that two infected laboratory employees, independently of each other, went straight from the WIV to the wet market, perhaps a week apart. Without infecting any of their colleagues. Without infecting anyone in their families. Without infecting friends. Without infecting anyone on the bus or train. Without infecting anyone at the store where they shop. But they both infect people at the wet market, and only there. They then go home again. Again without infecting anyone else. They go to work without symptoms. Every day for a couple of weeks. And then suddenly a superspreader event occurs at the wet market.
   
   Likely? Nope.
7. There is still no evidence that WIV had viruses similar enough to have been the “strain” of SARS2. The closest virus identified there is RaTG13, which is only about 96% similar, and everyone agrees that it therefore cannot have been used to create SARS-CoV-2. That would require a virus that is more than 99% similar.
   
   The closest virus found in nature is some variants of the BANAL viruses in Laos, but even these are not similar enough to be used as a strain to create SARS2. We know this because overviews of the viruses they were researching at the WIV were published several times, including in 2018 when they published a list of all the viruses they had. Only ten of them were sarbekoviruses, and none of them were close to SARS2.
   
   The DEFUSE application also listed 180 coronaviruses, and a study submitted for publication in 2019 listed 200 coronaviruses in the collection, without describing any virus resembling SARS2. And remember, before the pandemic, there was no reason for them to keep such a virus hidden.
   
   DEFUSE (which was never funded and therefore unlikely to have been carried out) also based its application on the use of the WIV1 virus, which is only 80% similar to SARS2 and therefore cannot have been the source.
8. No signs of genetic manipulation of the SARS2 virus. The furin cleavage site is inserted “out-of-frame” and is not a very ‘good’ furin cleavage site.
   
   Researchers would not have “inserted” such a suboptimal furin cleavage site. They would have used an existing furin cleavage site that they knew was effective, as has been done in previous experiments, rather than creating something completely new that they had no idea would work.
   
   And simply inserting a furin cleavage site is not enough in itself, because the SARS2 furin cleavage site has properties that optimize transmission that we did not know about before 2020. (I write about all this in more detail in a previous blog post.)
9. Two of the three species in which SARS1 was found, raccoon dogs and civets, were found in wet markets, which is where most positive environmental samples of SARS2 were found.
   
   It is worth mentioning that earlier in 2019, three market stalls were fined by the Chinese authorities for illegally selling animals. Two of these stalls tested positive for SARS2. Coincidence?
10. An unexpectedly large proportion of the very first cases of infection recorded even before the wet market came under suspicion were nevertheless linked to the wet market. As these cases were recorded before the wet market was suspected, this cannot be the result of confirmation bias.
11. In samples from other markets in Wuhan, almost no positive environmental samples with SARS2 were found. This makes no sense if it was humans who brought the infection to the markets, rather than it coming from animals there. Only at the Huanan wet market were there massive positive traces of SARS2.
    
    We also know the number of samples taken and that most positive samples were found in the southwestern corner where civets, palm civets, and bamboo rats were kept in cages. Ergo, this pattern cannot be due to “sampling bias.”
    
    Incidentally, some positive samples were also found in warehouses that supplied the wet market with animals, which is also a fairly clear indication of a link between SARS2 and the animals in question as intermediate hosts.
12. SARS2 is a recombinant virus, and all the building blocks required have already been found in nature in various coronaviruses.
    
    It is highly plausible that SARS2 emerged from among the billions of virus experiments through random recombination that occur in billions of viruses in billions of animals every single day. It is significantly more plausible than them having stumbled upon this perfect combination of the furin cleavage site, the receptor binding domain and the other components that required knowledge that was not available in 2019, purely by chance, within a few months of research on a few generations of viruses.
13. A virus cultivated in a laboratory in, for example, humanized mice would not be particularly infectious in humans from the outset. It would, however, be highly infectious in mice, but that was not the Wuhan variant of SARS2.
14. No evidence of mysterious events in Wuhan in the fall of 2019.
15. No WIV employees reported infection, mysterious illness, or anything else unusual happening there in the fall of 2019.
16. No documented COVID-19 cases in Wuhan before December 2019.
17. Historical precedent that the same animals as at the seafood market, at the same distances, with the same source (bats) have given us coronavirus outbreaks in humans in the past (SARS1/MERS).
18. No evidence of dangerous safety conditions at WIV.
19. No timeline or hypothesis for a lab leak has been presented that does not contradict scientific evidence or is internally inconsistent.
20. No evidence that the six miners who fell ill in the Mojiang mine in 2012 had SARS2. Analyses suggest that this was not the case.
21. SARS2 could infect humans, but was not optimized for human infection. The furin cleavage site could have been better, and it became more infectious and more dangerous after further evolution later in the pandemic.
22. Based on what we know about the World Military Games, this cannot have been a superspreader event in October 2019.
23. A majority of US intelligence agencies believe zoonosis is most likely, and they all agree that the virus does not show signs of genetic manipulation and that there is no evidence of GoF research at the WIV.
24. The three WIV employees who were allegedly ill in November 2019 do not fit Bratlie's timeline (virus database taken down in September, World Military Games and mysterious events in October), and they deny having been ill. They did not even work on live virus research, some had symptoms that did not match COVID-19, and tested negative for SARS2 in January 2020.
25. Criticism of the key studies arguing for the wet market as the epicenter only criticizes a couple of isolated analyses and does not find that an epicenter was more likely to be closer to the WIV or far away from the wet market.
26. Previous lab leaks have shown clear links between infection and the laboratory, and have not involved new viruses, only known viruses that could infect humans.
27. SARS2 looks like “a jigsaw puzzle”, which is exactly what you would expect from a recombinant sarbecovirus. There are no signs of genetic manipulation, quite the contrary.
28. The reproduction rate inside and outside the wet market suggests that there was no superspreader event there, but that the infection originated there.
29. Both similar furin cleavage sites and almost identical RBDs are found in other coronaviruses in nature.

Arguments for lab leak

After reading the book, some reviewers have become more convinced that a lab leak is likely. I did not. On the contrary, it became even clearer to me how this hypothesis has no coherent, plausible narrative to stand on, but also how dishonest the argumentation is, and how extremely many “what if” arguments it is based on.

Is there anything that makes me still consider the lab leak hypothesis plausible?

Well, here are the points that I believe still prevent zoonosis from being 100% certain:

1. We have not found either the source of SARS2, i.e. the natural reservoir, or the intermediate hosts. Until we find this, lab leak can never be ruled out with 100% certainty. And even if we find them, there is always a small possibility that lab leak was the cause.
   
   But remember that we were not expected to have found these yet. Historically, we know that it can take a very long time, so the lack of these does not argue for a lab leak either.
2. The hearings in the US revealed that there was probably not full transparency about what kind of research was going on at the WIV. However, there is no evidence that the virus was created in a laboratory, given that no virus candidate that could have been used as a virus strain has been found, but this does not completely rule out the possibility that research that could in theory have created SARS-2 took place there.
   
   But again, the pattern of infection spreading from the seafood market, the lack of evidence of infection cases at the lab, no signs that anyone there knew of any infection incidents, published virus overviews that show no similar virus in the WIV before 2019, etc., nevertheless significantly weaken this point.
3. Researcher Yusen Zhou, who had collaborated with Shi Zhengli on vaccine research, applied for a patent for a COVID vaccine in February 2020. Bratlie writes that it was the “very first vaccine,” which is not true, but it is still early. Others had applied for patents for vaccines before him, but he was the only one who had conducted such extensive animal studies as he described. Some experts believe that it would have taken at least 3-4 months to complete these studies, so he must have known about SARS2 by November 2019 at the latest, several weeks before the virus was sequenced and publicly described.
   
   At the same time, he was one of the leading experts in the world on this, with extensive experience with this type of virus, so others believe that it is not entirely unlikely that he could have achieved this in just a few weeks, assuming that all the experiments went according to plan. However, most experts do not seem to attach much importance to this, because it can be explained naturally.

These are the only three points that remain for me after reading the book that open up the possibility of a lab leak. None of them carry much weight as evidence of a lab leak, but they mean that a lab leak still cannot be ruled out, even though I believe that the evidence for zoonosis makes me at least 90% certain that this is the correct scenario.

Bratlie's hypothesis

Towards the end of the book, Sigrid Bratlie presents her own hypothesis about what caused the pandemic. Here is a review of it, where her points are summarized/paraphrased in highlighted text, with brief comments under each point:

1. Zhengli Shi and Peter Daszak found a virus in the Mojiang mine that they named BtCov/4991.
   
   True.
2. Miners in Mojiang fell ill and half died from what may have been COVID. They used the virus they found there to insert a furin cleavage site in research at the WIV in order to create vaccines against such a potential virus, which enabled Yusen Zhou to be so early with vaccine patents.
   
   The problem with this is that the miners were unlikely to have been sick from a virus, but rather from mold. And if it was a virus, there is no evidence that it was anything that could have been SARS2. There is no evidence that WIV had any virus more closely related to SARS2 than RaTG13, which is not similar enough to be used.
3. An infection accident occurred at the WIV, and measures were taken to limit the damage in October 2019.
   
   There is no evidence for this. There are no plausible signs that anything special happened at the WIV in October 2019. And if something like that did happen, it conflicts with Bratlie's claims of infection in Europe much earlier.
4. The first cases of infection showed no symptoms or mild illness, and more serious cases were interpreted as influenza and went under the radar.
   
   This conflicts with the claim that the virus originated in the Mojiang mine, because there the virus was 50% fatal. Did they do RoF research - Reduction of Function? Why then does Bratlie write so much about GoF?
5. The World Military Games in Wuhan in October 2019 were a super-spreader event that spread the infection to other parts of the world.
   
   No data supports this. On the contrary, there is no evidence that participants there started any epidemics in their home countries. Virus analyses also do not support this argument.
6. The virus evolved further, first to the A lineage, which later became the B lineage. A person infected with the B lineage went to the Huanan wet market and started the pandemic.
   
   Genetic analyses argue against B evolving from A in humans. They appear to have jumped from animals to humans in two separate events, both in wet markets.
   
   This point also contradicts one of her central claims that the virus was so well adapted to humans from the outset. Here, however, she writes that the virus evolved in humans for almost two months before it was potent enough to start a pandemic. This contradicts the idea of a virus genetically modified to be particularly infectious in humans.
   
   We also know that the wet market was not a super-spreader event, and it is extremely unlikely that the pandemic would have started there – of all places in Wuhan – if it had been brought to the market by a human.
7. Once the infection was discovered, testing was concentrated around the market, which led to a skewed picture of the spread of infection. The Chinese authorities removed the animals and took down the virus database.
   
   The virus database was taken down in September, so that point strongly argues against measures to conceal the infection only being implemented after the infection was discovered in December.
   
   It is also incorrect that data was primarily collected on patients with links to the market. Analyses where all these have been removed still show an epicenter at the seafood market. And there is patient data from before this link to the seafood market was suspected, and it also points to the seafood market.
8. The Chinese were slow to alert the world, but when the virus genome was published in January 2020, researchers discovered that it appeared to have been genetically manipulated. It also appeared to be well adapted to human infection.
   
   Again, this contradicts her argument that the virus had already been circulating for a long time in humans, with ample opportunity to adapt. We also know that the virus became much more contagious in humans only later in 2020, so it was by no means optimized for humans. Bratlie herself says that the virus spread to other countries in early fall 2019, but none of these cases led to infection in others, which also argues against it being so well adapted to humans.
   
   The reason researchers suspected genetic manipulation was that it had an RBD adapted to humans and a furin cleavage site. After a short time, it was found that both of these already exist in similar viruses in nature, and the idea that only genetic manipulation could have created such a virus was therefore abandoned.
9. Shi Zhengli tried to cover up anything that could point to her as the creator of the virus. Bratlie does not believe that the published genome of RaTG13 is correct and believes that Shi has manipulated data to hide the fact that it may have been used as a virus strain for SARS2.
   
   There is no evidence for any of this. It is just speculation based on rumors and data that is misrepresented throughout the book.
10. Key researchers such as Kristian Andersen and Ralph Baric warned the authorities and the research community that this could be a lab leak, but it could cause so much trouble that they all got together to cover up this conclusion and lie about what the data showed.
    
    There is no evidence for any of this. It is just speculation based on rumors and data that is misrepresented throughout the book.