"Oh my god, there’s been an outbreak of chocolate-y goodness near Hershey, Pennsylvania! What do you think happened?"
Or why I think SARS-CoV-2 was engineered in a lab.
Over the last few months, I’ve become convinced that SARS-CoV-2 was genetically modified at the Wuhan Institute for Virology, probably in a collaboration with American scientists. The point of this document is not to present any novel arguments, but mostly just to check myself that I actually understand the arguments that are out there. Most of the evidence I know comes from reading the tweets of Alex Washburne, Alina Chan, Richard Ebright, Gilles Demanauf, and the various groups who compiled and FOIAed the emails and slack conversation compiled below, as well as the anonymous twitterati with the handle @RAEMKA1 who made some of the amazing text graphics. I hope some people find this compilation useful. I also hope people will correct the parts that are wrong—which I am sure there are some of. Nobody should regard this as an authoritative source of information.
The Molecular Evidence
The main reason to suspect that SARS-CoV-2 is a laboratory construct is its unusual molecular structure. At this point, everyone has probably heard that the suspicious thing about the virus is its “furin cleavage site.” What is that? And what’s unusual about it? To be successful, a virus like flu or covid needs a spike protein to do two things. First, it has to bind to its receptor, then it needs to enter the cell via the receptor. For the spike protein to enter, however, it needs an assistant: specifically, something than can "cleave" the spike protein into two parts. The worst flus are "cleaved" by furin, an enzyme that is extremely common because it’s essential to Golgi bodies. So flus, and some others coronaviruses, have "furin cleavage sites" (FCS). These are points in the spike protein that furin can cleave to help activate this cycle. SARS-CoV-2 comes from a clade of bat corona viruses called sarbecoviruses. The sarbecovirus evolutionary tree we knew of pre-COVID spanned >80 genomes (with SARS-CoV-1 being just one of them) and >1,000 years of evolutionary time, and on that tree there wasn't a single FCS. The FCS on SARS-CoV-2 is on the S-spike between so called S1 and S2.
This is a nice video illustrating the process, except that it shows transmembrane protease serine 2, which cleaves the S spike at S2’. This needs to happen in addition to furin cleaving at S1/S2
So the question is: where did it come from? There are large number of anomalies that suggest it was put there by researchers at the Wuhan Institute of Virology (WIV). One minor reason to suspect this is that sarbecoviruses come from “Least Horseshoe” bats, which are not native to the area of Wuhan, but we know that researchers at WIV had been collecting these viruses for years. (I’ve seen what appear to be some exaggerated claims about just how far away the bats are, but they are definitely not swooping around the city of Wuhan.)
Of course, we all know the received view on how SARS-CoV-2 got to Wuhan (via wild animal trade, raccoon dogs, and the “wet market”--more on this below) so the most interesting facts have to do with the molecular structure of the FCS and surrounding RNA. FCS have long been an area of interest to virologists who study wildlife zoonosis. This is because there are a few well-documented examples of a FCS increasing the zoonotic ability of viruses. Influenza acquired a FCS during passage in chickens (this has happened one known time with influenza, despite it being monitored for decades). A bovine coronavirus acquired an FCS during passage (also n=1), and MERS-CoV had an FCS. This was heavily studied by researchers at NIAID. The mere presence of an FCS was alarming to top virologists Kristian Andersen and Eddie Holmes (who we will discuss later). They made this clear to scientist Sir Jeremy Farrar, who tells the story to us in his book “Spike”.
But given what we know today, it is not just that the FCS is there, it’s how odd it is. What is so odd about SARS-CoV-2’s FCS? There are three separate things (other than the already strange fact that the virus has an FCS), each of which on its own seems extremely improbable if the virus evolved naturally in bats. All three together are remarkably striking.
1. Everyone knows that DNA and RNA code for proteins with triplets of nucleotide base pairs coding for individual amino acids. What I never knew until I did a deep dive on Covid is that different organisms can code for the same amino acid with slightly different triplets. Not only that, but different organisms “prefer” different codings. If you use the coding for an amino acid from organism A in organism B, it can cause various problems, including slowing down the transcription process. The most common coding for the amino acid arginine *in humans* is CGG. But CGG is extremely rare in sarbecoviruses. SARS-CoV2 contains the codon pair CGG-CGG for two back-to-back arginines right in the FCS. When genetic engineers take a trait from one organism and try to put it into another (like taking the gene for making omega-3 fats from spinach and putting it in pigs) they usually “optimize” the codon, changing the coding for each amino acid into the one that the receiving organism prefers. Thus, it seems as if such a procedure occurred with SARS-CoV-2. It is true that some coronaviruses do (rarely) contain CGG, but two back-to-back and right in the suspected-to-be-engineered area is extremely unlikely.
2. The amino acid sequence of the SARS-CoV-2 FCS is identical to that of a peptide found in human cells: ENAC-α. The sequence is RRARSVAS. ENaC-ɑ is expressed in the nasal epithelial cells, type II alveolar cells of the lungs, tongue keratinocytes, and colon. Unsurprisingly, therefore, all of these comprise COVID-19 pathophysiology. Various people have debated the “odds” of this match between ENaC-ɑ and the FCS happening randomly. I won’t get into the weeds of this, but there are few things to keep in mind:
a. In a bat, (or pangolin, or racoon dog) there would be no selective pressure to evolve this sequence, so it would be a pure coincidence.
b. The coincidence isn’t, as some people have made it out to be, that the sequence matches any known sequence in the human genome. It’s that it matches a functional sequence, and one that is functional, moreover, in the respiratory system. SARS-CoV-2 binds to ACE2 receptors, which are everywhere in the human body, but ENAC-α is only in the nose, lungs, and colon. And it was known that genetic dysfunctions related to ENAC-α cause morbidity. These factors are almost surely why it was of interest to gain of function researchers at the University of North Carolina where DEFUSE collaborator Ralph Baric and his wife Tony Baric (both Principal Investigators of DEFUSE) worked.
3. SARS-CoV-2 has all the hallmarks of genetic engineering. In order to insert something like an FCS into an RNA coronavirus, you first need a full-length DNA clone. And to make a full-length DNA clone, you first need to assemble the full-length clone from smaller building blocks. The most common way to do this is to use special enzymatic cutting/pasting tools called "type II restriction enzymes". Using this technology, you could build any virus whose genome you received over email. SARS-CoV-2 is very rare in having unusually spaced type II restriction enzyme recognition (cutting/pasting) sites.[1] The mutations that "move" these sites around are exactly the silent mutations researchers use when making clones, and there is a significantly higher concentration of silent mutations within these sites than the rest of the genome (5e-8 P-value for covid’s likely ancestor: RaTG13).
That summarizes the molecular evidence. In the rest of this piece, I’ll summarize the criticisms of the evidence for the standard story (that the virus came from a wet market), review the evidence that the most knowledgeable central players in this debate believed it was a laboratory construct and lied about their beliefs to cover that up, and that they did so under the direction of very powerful scientific bureaucrats like Anthony Fauci, Francis Collins, and Peter Daszac, and review the evidence that researchers were trying to do exactly the kind of research that would best explain the 3 molecular anomalies I discussed above.
The “Wet Market” and Worobey et al. 2022.
Recall that sarbecoviruses come from bats that live quite far from Wuhan. We know that WIV researchers were out in the field collecting these, so even if SARS-CoV-2 is natural, one possible way it could have come to Wuhan is via this collection process (“natural lab escape” as opposed to “engineered lab escape”). The other candidate explanation, the one that is preferred by the bien-pensant, is that it came to Wuhan via racoon dogs that were for sale in the Huanan Seafood Market, aka the “wet market”. This claim was said to have been established in Worobey et al. 2022. (https://www.science.org/doi/10.1126/science.abp8715) This is the paper titled “The Huanan Seafood Wholesale Market in Wuhan was the early epicenter of the COVID-19 pandemic” that many people think conclusively shows that we know where the pandemic started from. There are two problems with this paper. The first is the model, the second is the data the model relies on. We’ll start with the model. The inference method involves geospatial analysis that relies on a “centering model”. The idea is simple: “We hypothesized that if the Huanan market were the epicenter of the pandemic, then early cases should fall not just unexpectedly near to it but should also be unexpectedly centered on it” (Woroby et al. 2022). A centering model assumes that most early cases are infected near the origin, and that the residences of the people who are infected are distributed in a circular pattern around that origin. But Daniel Walker has pointed out that is not how humans move and infect each other in a dense urban environment with a complex transportation network. There are many examples of this. Walker shows that if you look at how Covid-19 spread through Singapore when it arrived there, it very quickly violated the centering assumption.[2] Worobey et al. were never given access to any real data from China. They had only a very small subset of the cases. The problem is that as the number of cases goes up, as the Singapore dataset shows, the centering model gets more and more brittle.
The problem with the number of cases is much larger than that, however. First of all, we know that the number of identified cases is much smaller than the total number of people who would have been infected. The authors of “The Huanan Seafood …” insist that that the data they have is representative. They deny, in other words, that there was “ascertainment” bias in collecting the cases. Ascertainment bias is what you would have if you looked for discarded bottlecaps in the dark, and concluded that they were primarily discarded in the vicinity of lamp posts. But recently, George Gao, director of the Chinese Center for Disease Control, admitted that there almost certainly was ascertainment bias in the collection of cases. He says there was special focus on the Seafood Market and people associated with it in the early search for cases.[3] In fact, Worobey et al. claim to use their model to show that there is no ascertainment bias. So, given that we now know there was ascertainment bias, what does this show about their model? Finally, Gilles Demaneuf has conclusively shown that that the data Worobey et al. had access to were only a subset of the cases the Chinese authorities are aware of. And we have no idea what selection criteria they used in making that subset available.[4]
It has also been shows that someone deleted some of the early sequence data from the NIH’s “Sequence Read Archive”. Some of this data has been recovered and suggests that there were cases with a closer ancestry to bat viruses than any of the Seafood Market cases. [5]
“So friggin’ likely”: “Proximal Origins” and the effort to manufacture consensus of natural origin.
In discussing Worobey et al., moreover, I would be remiss if I did not engage in a bit of ad hominem. Four of Worobey’s co-authors are authors of the now infamous 2020 Nature Medicine paper entitled “The proximal origin of SARS-CoV-2”. The five authors of that piece are Kristian G. Andersen, Andrew Rambaut, W. Ian Lipkin, Edward C. Holmes and Robert F. Garry. Of those five, only Lipkin is not an author of “The Huanan Seafood..” The reason this is significant is that we now know that the “Proximal origins” paper was written in bad faith. We know that the paper was written in bad faith because it was submitted on Feb 4th, and concluded that “Our analyses clearly show that SARS-CoV-2 is not a laboratory construct or a purposefully manipulated virus,” but just three days earlier the lead author had privately shared with his co-authors that ““I think the main thing still in my mind is that the lab escape version of this is so friggin’ likely to have happened because they were already doing this type of work and the molecular data is fully consistent with that scenario.”
Indeed, all five co-authors, just days before the paper was submitted, shared a number of similar sentiments that have been compiled by various sources on Twitter. The very best are by someone named @RAEMKA1.
Proximal Origins contains virtually no arguments. We know that it was published only because of the influence of the authors and their “mentors”: Francis Collins, Ron Fourchier, Anthony Fauci, Peter Daszak and Jeremy Farrar. I call them their “mentors” because we know from emails, Slack messages, and other sources that the article was written under the direction of Farrar, who served as unofficial liaison between the virologists and Fauci and Collins, and that all of them pushed hard for the draft that was submitted on February 4th. Farrar wanted a “neutral, respected scientific group” to say that a credible group of scientists should rein in the discussion about Wuhan’s labs “before that debate gets out of hand with potentially hugely damaging ramifications.” And we know that Farrar, then head of the Welcome Trust and now chief scientist of the WHO, “pushed” Nature to publish because he wanted to “counteract rumours of bioweaponeering.”
After a phone conversation with Fauci, Holmes referred to what they were being asked to do as a “Big ask!”, to which Andersen replied “Destroy the world based on sequencing data, yay or nay? Makes sense what he’s saying but man that’s hard to pull off.” If, as many suspect, what Andersen thought would be hard to pull off was to manufacture a consensus around the wet market theory, then I find his lack of faith disturbing. They pulled it off quite well.
Another terrific graphic by @RAEMKA1 captures the back and forth between the authors and various interested parties.
Because of some of these shenanigans, many people have called for the retraction of the “Proximal Origins” paper. Different people’s arguments for this vary, but mine is that it was published as expert opinion, but did not actually reflect the opinion of the experts who put their names on it. Rather it reflected the political pressures they were put under. Indeed, the official view of the journal is that it is not retractable because it was “merely an expression of opinion.” Perhaps this is right, but it was not taken this way either by the popular press or by many very powerful government and international agencies.
We also should discuss the February 19, 2020, letter in The Lancet titled “Statement in support of the scientists, public health professionals, and medical professionals of China combatting COVID-19,” written by a group of 27 researchers from medicine, biology, virology, epidemiology, and public health. They wrote:
Scientists from multiple countries have published and analysed genomes of the causative agent, severe acute respiratory syndrome coronavirus (SARS-CoV-2), and they overwhelmingly conclude that this coronavirus originated in wildlife, as have so many other emerging pathogens. This is further supported by a letter from the presidents of the US National Academies of Science, Engineering, and Medicine and by the scientific communities they represent. Conspiracy theories do nothing but create fear, rumors, and prejudice that jeopardise our global collaboration in the fight against this virus. We support the call from the Director-General of WHO to promote scientific evidence and unity over misinformation and conjecture.
When this letter was cross-posted on Change.org, it received over 20,000 signatures, many of them by people representing themselves as relevant domain experts. We now know the letter was orchestrated by fifth author Peter Daszak, the president of EcoHealth Alliance, which has millions of dollars of funding for the kind of research that might be called into question if the lab-leak hypothesis were accepted. He wrote privately that he wanted the letter to “avoid the appearance of a political statement.”
Daszak signed a statement declaring he had no conflict of interest. This led The Lancet to take the unusual step of publishing an “Addendum: competing interests and the origins of SARS-CoV-2” in which it discussed Daszak’s competing interests. Since that time, we have learned a great deal about Peter Daszak, and his organization’s role in sponsoring the exact kind of research that very easily could have led to the creation of SARS-CoV-2. Don’t take my word for this, consider the words of Kristian Andersen and Andrew Rambault, two of the authors of Proximal Origins. (For context, PREDICT was the joint work of Daszak and Baric.)
“One thing I find kinda funny here – all of this work getting bat samples was supported by PREDICT. So if they’re not able to predict the pandemics they themselves cause, then I’d say their program is in pretty bad shape,” Andersen said on February 3.
“Perhaps they had planned a press conference predicting which virus would cause the next pandemic but then it escaped from the lab early?” Rambaut said.
“Any [intermediate] animal being identified would be beneficial to them – otherwise we’re all going to point fingers at them telling people that they’re so shit that they can’t even predict the outbreaks of their own making…Too harsh?” Andersen said on February 7.
It’s really too bad these guys became paid liars because, before they did, they were quite funny.
DEFUSE, and the evidence that American and Chinese researchers were collaborating to do everything that it looks like was done to make SARS-CoV-2.
In this last section I review the evidence that researchers were trying to do exactly the kind of research that would best explain the presence of an FCS in SARS-CoV-2 and its molecular anomalies that I discussed above.
The strongest evidence comes from a 2018 grant proposal to DARPA called “Project Defuse: Defusing the threat of bat-borne coronaviruses.” Authors of the proposal include leader Daszak, as well as UNC scientist Ralph Baric, and Wuhan Institute of Virology senior scientist Zhengli Shi (who Scientific American referred to as “Bat Woman”). The proposal was to, wait for it, synthesize spike proteins with an FCS and insert them into coronaviruses from the same clade as SARS-CoV-1. (And keep in mind that Baric was affiliated with a group that had an interest in studying ENAC-α as a known furin-cleavable peptide with a presence in the respiratory system.)
To be absolutely clear about this, DARPA rejected the application. But many suspect the work was carried out anyway. We know that Daszak had the ability to carry out the work even without DARPA funding. We know this because of an email he sent on April 28, 2020. The email concerned a different project called “Understanding the Risk of Bat Coronavirus Emergence” another collaboration between EcoHealth, Baric at UNC, and the WIV. It was funded by the NIH but suspended in April 2020. Daszak wrote “My plan is to continue this work, unfunded for now…” So, it’s not at all unlikely that DEFUSE was carried out despite the DARPA rejection.
Everything we know about DEFUSE is alarming. Formally, the grant proposes to do all the spike protein work at UNC, but a draft comment on a MS word document reads “Ralph, Zhengli. If we win this contract, I do not propose that all of this work will necessarily be conducted by Ralph, but I do want to stress the US side of this proposal so that DARPA are comfortable with our team. Once we get the funds, we can then allocate who does what exact work, and I believe that a lot of these assays can be done in Wuhan as well…”
In another comment, he wrote “I’m planning to use my resume and Ralph’s. Linfa/Zhengli, I realize your resumes are also very impressive, but I’m trying to downplay the non-US focus of this proposal so that DARPA doesn’t see this as a negative.”
There are two reasons this is significant. First, Daszak publicly tried to refute the claim that DEFUSE might have been at the root of the Covid-19 pandemic by reminding people that the work was to be done at UNC, not in Wuhan.[6] But that was before he knew that his draft comments would be FOIA-ed and that we would come to know that was not in fact the plan. The second reason is that part of the reason Daszak wanted the work to be done at WIV was that he knew it could be done there more cheaply under BSV-2 level biosafety, whereas at UNC it would be expected to be at least BSV-3. (It’s a four-level system.) In comments, he [edit: this comment is Ralph Baric’s, not PD’s] wrote that US researchers would “freak out” about this kind of work being done under BSV-2.
[1] https://www.biorxiv.org/content/biorxiv/early/2022/10/20/2022.10.18.512756.full.pdf
[2][2]
https://twitter.com/danwalker9999/status/1595653966054187009
[3] https://www.bbc.co.uk/sounds/play/m001ng7c (Begin listening around 24:30 if you want to skip to the relevant comment.) Alina Chan has highlighted this fact.
[4] https://www.washingtonpost.com/opinions/2022/11/17/covid-early-cases-wuhan-china-mystery/
[5] https://www.biorxiv.org/content/10.1101/2021.06.18.449051v1
[6] https://www.science.org/content/article/science-takes-back-seat-politics-first-house-hearing-origin-covid-19-pandemic
The fact that you managed to learn something from twitter--and enough to write a detailed and convincing essay--is impressive!
Nice compendium of (parts of) the case for lab origin. You left out the sort of Bayesian no-miracles line of reasoning (what a coincidence, same city as WIV! etc.), except in your title, was that just to focus on the easiest-to-appreciate stuff? Another Bayesian-type argument: why did the Chinese absolutely *freak out* about C19, for years on end? Easy to understand if they knew it was a GOF human construct designed for heightened transmissibility and lethality.
You could also add the fact that some government investigators were literally paid to change their conclusions to favor natural origin instead of lab leak. (I can look up this episode if you don't have it at hand.)