You know how it feels when you know someone’s bullshitting you?
That’s how it felt to be a researcher in pathogen spillover reading Proximal Origins.
I don’t say “bullshitting” lightly. There are many papers whose conclusions I don’t agree with, often because their methods are limited and there are other explanations. For those papers, I typically say that I don’t agree with their conclusions, and then I explain why. Most papers in science make pretty limited conclusions, so it’s rare that I’m actually upset about a paper. However, when I first read Proximal Origin I was viscerally upset, increasingly so the farther into the paper I went. At the beginning of the paper, I had high expectations that a group of rock-star authors would deliver a rock-star paper, but by the end of the paper I felt lied to.
Now, Slack communications by the authors of Proximal Origins have been released and it’s clear: they didn’t actually believe their claims that SARS-CoV-2 was a zoonotic virus and that a lab origin was implausible. In fact, the first author Kristian Andersen said it best while writing the paper: a lab origin is “so friggin’ likely”. While the authors were privately being Slack Bros confessing that SARS-CoV-2 most likely came from a lab, a consequence of Peter Daszak’s PREDICT efforts to collect coronaviruses and ship them to Wuhan, they were saying to journalists, US federal officials, and the public outside their circle of bro’s that a lab origin was “debunked”, “implausible”, “illogical”, etc.
I’m happy everyone is now seeing the Proximal Origin paper for what it is. It was a propaganda piece, it was a lie. Almost a year before the Slack communications were revealed, I had finished the hard work of forecasting outbreaks and had time to catch up on the SARS-CoV-2 origins literature:
I’ve put my head to the task of independently examining the origin of SARS-CoV-2 and I’m discovering a highly unusual trail of literature from leaders in the spillover community, a trail of literature that reads to me like an effort to discourage investigations and tamp-down any questions of possible lab origin as a “conspiracy theory”.
Now, we have the receipts. Andersen et al. were concerned about the political implications of a lab leak, they thought if anyone serious accused China of making SARS-CoV-2 it would be a “shit show” and they conspired to lie about having ruled out a lab origin before it “creates more formal investigations”. Indeed, it was an effort to “discourage investigations and tamp-down any questions of possible lab origin as a ‘conspiracy theory’”.
How did I and others know that in September 2022? Personally, I try my best to limit my words to claims I can justify with evidence - that’s the skill of being a scientist - and so instinctively, intuitively, my brain had amassed enough evidence to accuse colleagues of research misconduct; it was shockingly obvious to me then and it’s shockingly obvious to everyone right now. I want to explain why it was shockingly obvious to me then because, right now, there are other things that are shockingly obvious to me but currently not appreciated by everyone. I hope that by explaining this process I can help everyone gain new tools for identifying bullshit even if it’s from scientists.
“Simulations say Spike protein binding is not ideal so SARS-CoV-2 isn’t from a lab”
Bullshit.
This is like drawing a stick figure in the dirt, saying the stick figure is not ideal for some simulation and so therefore it wasn’t drawn by a person. You weren’t drawing the stick figure to make it ideal for a simulation, you were making it for other purposes.
Scientists in the lab don’t ticker with things to make them “optimal” for any computer simulation, and with computer simulations something that is 99.999% “perfect” is still not “optimal” compared to something 100% “perfect” so the notion of “optimal” is a straw man. This straw man overlooks a critical detail: the SARS-CoV-2 spike protein is better at binding humans than it is at binding other animals, including bats. Under a natural origin, we would expect the protein is better at binding bats or its intermediate host than it is at binding humans.
Even if the protein were better at binding bats than humans, scientists work with bat viruses in the labs all the time. For example, a bunch of researchers in the Wuhan Institute of Virology modified restriction sites to make an infectious clone of a bat SARS-CoV virus they named rWIV1. The researchers then took other bat SARS CoVs and inserted their spike proteins into rWIV1. Any one of these chimeras would have “suboptimal” binding of the human ACE2 receptor, yet all of them are indisputably from a lab.
Saying “Human ACE2 binding is suboptimal therefore SARS-CoV-2 couldn’t have come from a lab” is bullshit, and this bullshit is obvious to anyone with an undergraduate understanding in biology who knows about the myriad objectives of lab work, and how none of them are to optimize binding in a computer simulation. The authors all have at least an undergraduate understanding in biology, so their bullshit here felt like a lie.
“If genetic manipulation had been performed, one of the previous backbones would have been used”
This is absurd. Every backbone ever made first existed in an unpublished state. If the rWIV1 example from 2016 had caused a pandemic, it would indisputably have been a laboratory-related accident with a genetically engineered virus that was not one of the previously published backbones.
This argument is so obvious. Everything that exists today was new at some point in time, and scientists are constantly making new things. Insofar as labs make new viruses, which they do, not every laboratory accident will be from an old virus. We already had a SARS-CoV backbone, but researchers found somewhat distant bat SARS CoVs and so they made a new backbone, rWIV1, to satisfy their research curiosities of making chimeric bat SARS CoVs.
By this point in Proximal Origins, you have every right to be livid. There is no way the authors believe this. They are making outlandish arguments that are so clearly not true. They never give a fair defense of a laboratory-related origin and such a strong argumentative bias is not random. Bullshit.
Three possibilities: either A, B, or C (ignore D, E, F)
The authors present three possibilities: a natural origin with selection prior to “the jump”, a natural origin with selection after “the jump”, or selection during serial passage.
The authors give not-terrible hypotheses for how natural origin with selection prior to the jump could occur, they talk about possible hosts, but, because they called the spike protein “suboptimal”, they didn’t contend with the fact that it binds humans better than bats. The authors give a similar alright-ish explanation of a hypothesis the virus jumped into humans and acquired the furin cleavage site in humans.
The authors’ defense of the one and only laboratory scenario, meanwhile, is bullshit, both because of the weak defense and because they fail to consider other laboratory based scenarios. The neglect the possibility that a laboratory-based scenario could have found the SARS-CoV with a furin cleavage site in bat or pangolin CoV and then infected a lab worker, they neglect the possibility that the furin cleavage site was inserted with the glaringly obvious (BsaXI) restriction site surrounding the furin cleavage site. Then, towards the end, the authors say the most outrageous part of this section that is more or less: “this could have happened by serial passage in humanized mice, but that hasn’t been described”
Again, anyone who does research knows that a novel research product is not described for a long time until it is finally published. This is like saying the brand new Subaru next to the Subaru factory couldn’t have come from the Subaru factory because it didn’t have a “For Sale” sign on it (yet!).
“We observed ALL notable SARS-CoV-2 features”
The use of “ALL notable” features is bullshit.
The genome of SARS-CoV-2 has over 30,000 base pairs and over 1,000 mutations from the closest relative available to SARS-CoV-2. The authors looked at only 2 features: mutations in the receptor binding domain and the Furin Cleavage Site (FCS). The authors did not examine all the other mutations, they didn’t look at the codon biases, nor did they examine critically important sites for bioengineering: restriction sites.
The authors also didn’t examine important pieces of epidemiological evidence about the emergence of SARS-CoV-2. SARS-CoV-2 is a bat SARS CoV that emerged in Wuhan, far from the hotspots of wildlife SARS CoVs but the same city with a lab that had the world’s largest repository of… you guessed it… bat SARS CoVs. SARS-CoV-2 was highly infectious in the very first observed patients, capable of causing superspreading events and causing cases to double every 2 days in major early outbreaks. That high transmissibility is an unusual feature of SARS-CoV-2. Additionally, SARS-CoV-2 binds more efficiently to human receptors than to bat receptors. That’s an unusual feature of a bat virus.
Clearly, it’s an exaggeration for the authors to claim they observed “all” notable SARS-CoV-2 features. This language raised a red flag because the not only did they not examine ALL features, who are the authors to be sole determinants of what constitutes a “notable” feature? I think the Wuhan emergence, the high transmissibility, the restriction sites, and the more efficient binding of human than bat receptors are all notable features that the authors clearly did not observe.
“We do not believe that ANY type of laboratory based scenario is possible”
Similar to the bullshit above, the use of ANY type of lab scenario is bullshit.
Suppose the virus was completely unaltered in any way from a virus found in a bat, but that unaltered virus was injected into a humanized mouse and infected a researcher. The authors could not rule out that laboratory based scenario because it would be impossible to tell based solely on the genome of the virus.
More scientific data could swing the balance of evidence to favor one hypothesis or another
The authors’ last paragraph started off balanced, but, when describing evidence that could swing the balance of evidence to favor ‘one hypothesis or another’, the authors list only evidence that would swing the balance in favor of a zoonotic origin. They don’t mention that finding a close relative in an animal without a furin cleavage site may indicate a laboratory insertion, nor do they mention that research notebooks from laboratories could reveal a laboratory origin, nor do they admit that there could be other features in the genome which might reveal a laboratory origin.
I’m hard of hearing, and it’s obvious. If there’s a sound that startles everyone, I often don’t even budge. If sirens go off outside the window in the middle of the night, I sleep undisturbed. As surely as you can tell that I am deaf, I could tell the authors were blind to the possibility of a lab origin, other notable features in the virus, the circumstances of its emergence, and the evidence that could swing the balance to favor a laboratory origin.
These authors are not idiots, so I know they are not blind to such obvious limitations of their work. The authors’ previous work rarely issued such bold proclamations and ignored such obvious pieces of evidence. Consequently, this read like the authors were lying, like they were willfully ignoring what is obvious to people in the field.
What else is obvious? A barrage of bullshit
Most papers will have one or two sentences that make us shake our heads back & forth, think critically, maybe disagree, but then move on having slightly discounted a strong or unfounded claim.
Proximal Origin was not most papers. Instead, Proximal Origin made a series of outrageous claims and used a series of outrageous straw men for laboratory origin scenarios, every single one of which an undergraduate in biology should be able to identify as wrong. Suppose there’s a 5% chance that 5 unbiased experts in virology all make the same dumb mistake in a paper of such high consequence such that the mistaken claim gets through to the final revision. The probability of all 6 jaw-droppingly-ludicrous mistakes above occurring is 1 in 64 million, assuming the authors are unbiased. If we turned this into a hypothesis test of whether or not the authors are unbiased, we get a P-value of P=1.5e-8, so we’d confidently reject the hypothesis the authors are unbiased.
The authors didn’t stop at the Proximal Origin. Instead, they continued on their crusade of BS leaving a trail of jaw-droppingly-ludicrous claims and conclusions everywhere they went. They had a convoluted Midas Touch where every scientific paper they touched turned to BS.
Worobey et al. claimed to have found “dispositive” evidence or a lab origin (strong wording, like ALL notable, ANY), yet Worobey et al. failed to control for strong spatial biases in how samples were collected, both human samples in Wuhan and environmental samples inside the wet market.
Pekar et al. claimed that a simulation of HIV transmission is unlikely to generate a tree specifically like SARS-CoV-2’s evolutionary tree with two large branches at the base of the tree, so therefore two spillovers happened. However, SARS-CoV-2 is not HIV, their model is wrong, superspreading can create two big branches, and why can’t two lab leaks happen (e.g. two or heck three index patients infected by the same mouse)?
Crits-Cristoph et al. claimed in blaring headlines in The Atlantic that they found “the strongest evidence yet” (literally “strong” wording) of a zoonotic origin… but their analysis ended up just cherry-picking a single sample with animal DNA in an animal market, incidentally the sample that had the least SARS-CoV-2 RNA of any sample with SARS-CoV-2 RNA, so little RNA that we can’t confirm the RNA is even from that sample!
Yet, everywhere these authors go to create a novel monstrosity in the scientific literature, a fanfare of media has followed blaring headlines full of overconfident language completely unsupported by the arguments & analyses of the papers, and completely unresponsive to - and unwilling to amplify - critiques that correctly identify the flaws. We can articulate fatal flaws in their papers with single sentences, yet the authors never incorporate this feedback.
When we look at the barrage of bullshit we’ve received the past few years, and the propagandic behavior of the media and the authors, it’s not a surprise at all for me to learn the authors were lying. I’m not alone. Many other scientists could see through this veil of propagandic deception. The most frustrating part of this experience has been hearing other scientists uncritically repeat the BS they heard in headlines right back at us, and for us to tell them the reasons and logic why their arguments were unsupported or untrue… and for them to call us “conspiracy theorists”. We know those arguments, we read the over-hyped papers and saw the over-confident headlines, we provided reasons why they are incorrect and yet all the marionettes (h/t Richard Ebright) can say back to us is the same things they hear on the speakers.
We did not merely wander out, we are exiles of Plato’s cave. A narrative so potent has, like a sinister worm, consumed the brains of everyone inside the cave that they can’t even comprehend the existence of something else. They have no ears so they can’t hear our reasons, they have no eyes and so they are unable to see the graphs and diagrams we use to illustrate the right answers. As our voice gets hoarse and our eyes well up with tears in frustration, the worms instruct the subjects to turn and look at us with a rotten, toothless smile and call us “conspiracy theorist”. This narrative, and what it has done to scientific society, is grotesque. We aren’t just eroding the fabric of our shared reality, we are slowly bending the fundamental rules of these realities, the mathematics & logic, until they are incompatible, and we cannot even reason with one another.
When I say “ALL” I mean “ALL”. When I say “ANY” I mean “ANY”. If I were to say “dispositive” (which I have never said in my life of science before ridiculing the term), I would mean “dispositive”. It seems obvious that if I said “all apples are red” you would say “no, there are green apples”, but when the authors say they looked at “all the features” and we say “no, there are other features” it does not compute. It feels like we are talking with idiots.
Why?
However, these scientists are not idiots, and that’s the scary part. There is not a worm inside their brain; that would make it so much easier to understand and sympathize. Instead, there is something else happening. Because these scientists are not idiots, I hypothesized that maybe the scientists stared at the consequences of this discovery and considered a trolley problem, something to do with nuclear war and/or Trump’s reelection, and decided to intentionally lie to us, thinking it is for the greater good of all. However, by lying to us they are not letting the “all” have a vote in the matter, not only on the ethics of their decision but on whether their assessment of the consequences of a lab origin are correct. What if the authors thought “if we publish this, there may be a nuclear war”, but there exist ways unforeseen by the authors to almost certainly avoid nuclear war? What if the authors’ decisions to bottle up all the rage of a public so obviously deceived was actually more likely to cause nuclear war? The authors are not experts able to forecast the impacts of their actions on all of human history. I don’t know, and neither do the authors. It’s not their call, as scientists, to lie about science because of some perceived political impact. While we should be geopolitically aware, we must do so in a way that is not outright unethical and does not greatly undermine trust in science.
Reading their work, it was evident before - and it is proven now - the Proximal Origin authors sought to deceive the world. When we strip away the worm of deception, when we wash out all the BS of the literature, when we put our eyes, ears, and minds back to use, it’s clear that SARS-CoV-2 most likely came from a lab. This is what folk outside the cave have been shouting for quite some time now. The consequences of a lab origin may be severe. The authors of proximal origin were most likely persuaded to deceive by some mix of grant money and morals. My guess: Fauci, who ran US biodefense funding since its inception in 2002, probably told the guys that if SARS-CoV-2 is believed to be a bioweapon, it could lead to nuclear war. However, they may not have considered the possibility that SARS-CoV-2 is a laboratory accident from normal pre-COVID research, and that such an accident would not lead to nuclear war.
Their words on Slack provide some indication of their thinking at this critical juncture. During the call with Fauci, Holmes says “Big Ask!” and Andersen replies by asking “Destroy everything with sequence data, yay or nay?” On February 8th, Rambaut says they need to get the science out before rumors of a lab origin “create more formal investigations”.
The authors who have lied to us for years have not yet told us the whole truth about their call with Drs. Fauci, Farrar, and Collins.
I don't know the research backgrounds of the authors of the Proximal Origins propaganda piece, but isn't the most likely primary explanation that they believed some flavor of the idea that if the public became aware that the pandemic was the result of a lab accident in GOF research that all future funding for such (and maybe by extension, funding for all kinds of laboratory research) would be curtailed or eliminated?
Speaking as a biologist, I've always thought that GOF research was a mistake. Compared to the risk of accidental release - which, if you have even the tiniest glimmer of an intuitive understanding of statistics you'll quickly realize is a 100% probability over a humanly meaningful timeframe - what you learn from such research has absolutely trivial big picture value unless you are trying to build weapons. Which is categorically wrong.
I'm in favor of a global ban on GOF.
Through out this all there has never been any discussion of National Center for Medical Intelligence (NCMI). A defence intelligence agency responsible for tracking pandemics using spies and other means. They are the types that would break in to labs and get data, or climb into Chinese sewers. Our look at NSA data on Chineese cell phones to detect clusters at hospitals. They likely would have picked this up in October 2019 and fed it into the Presidential Daily Brief. Their org has never been mentioned throughout this entire episode. Which is odd.
If you were President and found out November 2019 there was an emerging virus, with unclear consequences, that may have spilled from joint US China research, what would you do?