A glowing green mouse was found:
The mouse was found outside a lab that proposed to make glowing green mice, but natural origin proponents would say that’s just a coincidence. There are labs everywhere, they say. They ignore the specific research of this lab, and they ignore that this village was previously used as the control group given the complete absence of wild mice in the village.
When we sequence the genome, we find the mouse has an unusual insertion compared to other rodents: a green fluorescent protein that doesn’t have an exact match in nature. There are, however, two other places that this broader feature of “fluorescent proteins” occur: in jellyfish, and in a grant proposal in the lab less than 1 mile from where the mouse was found.
When we look closely at the sequence in the mouse, we see that the amino acid sequence is very similar to that found in jellyfish, but the codons generating that sequence appear to reflect the codon biases of a mouse. Of course, this could also reflect codon biases of other unsampled viruses or organisms that contained the GFP, so this proves nothing, despite it being a well-documented tool and also proposed in that same grant proposing to insert the GFP in that same lab 1 mile from where this mouse was found.
The genome of the mouse also looks a lot like a clone made by methods commonly used prior to our discovery of the glowing green mouse. Folk say this proves nothing, that they would clone mice using different methods (neglecting the historical commonality of this method), and that, even so, everything is possible with evolution.
We ask the lab, “Is this your mouse?” and the authoritarian government orders its military to take control of the lab and refuse to release any databases or lab notebooks that could help us verify if this was their research product.
Some people say the government is “hiding animals”, but if this was their research product and they sampled more animals in the area, the odds of this animal being found in nature go down. By sampling more animals, we’ll have found more things in nature and not found another glowing green mouse, so the numerator would stay the same as the denominator grows.
Thus, our academic scientists would conclude that the mouse is a natural mouse. The location is just a coincidence when we neglect the grant proposing this work, the multiple features of the mouse matching exactly that proposed in the grant are also a coincidence as we are ignoring the grant, and the grant itself is just a coincidence because it didn’t get funded by one funder despite the researchers having other means to conduct that same inexpensive research.
Everything is indeed possible with evolution if we are sufficiently negligent in our handling of the facts and probabilities of the matter. In theory, it’s possible for a chicken to give birth to a hippo. It’s not very likely, but you’ll never find direct evidence disproving that the hippo, whose birth we didn’t observe, didn’t actually hatch from a chicken egg. A mouse with a jellyfish gene is possible. After all, 10% of our genomes are of viral origin, including enzymes in our saliva that we use to break down the sugar amylose, so there’s never direct evidence disproving the natural origin of the jellyfish gene, <hand-waving> possibly through the same mechanisms as our salivary amylase.
Such is the state of scientific affairs on SARS-CoV-2 origins. We have found a virus in an unusual location with multiple unusual genomic features all of which were specifically proposed in a grant by a lab in that exact location. Every single piece of strong evidence - the location, the furin cleavage site, the human-optimized codons, the restriction map of an infectious clone - is swept under the rug as natural origin proponents mislead the world and say there is a complete absence of direct evidence. They say the evidence suggesting a lab origin is nothing more than coincidences, and they refuse to estimate the probabilities of those coincidences under the two theories.
What a fantastical world these natural origin proponents live in! By their logic, by the unfalsifiability of natural origins of all things in nature not directly observed, a chicken egg can hatch into a hippopotamus and a mouse can acquire a codon-optimized jellyfish gene in the same city as a lab proposing to do exactly that. A bat SARS coronavirus can pop up in Wuhan next to the world’s biggest repository of bat SARS coronaviruses, a city typically used as a control group for bat SARS coronaviruses, and this bat SARS coronavirus can have human-optimized furin cleavage site, the restriction map of an infectious clone, and it could even have “Made In China” written in code in its poly-A tail, and it would all be a coincidence. Everything is natural in the fantastical world of natural origin proponents. I pity us wretches who can’t see the world through the lens of their unbounded credulity.
We’re left with the burden of rigorously estimating what is more likely. Did the hippo hatch from a chicken egg, or are we as scientists going to call bullshit? Did the mouse acquire a GFP from a jellyfish, or is that bullshit? Did SARS-Cov-2 emerge naturally in Wuhan with all the features of a WIV grant, with the Chinese military seizing control of the lab, refusing to share their databases and lab notebooks, refusing to allow independent audits?
Or is that bullshit?
I became a scientist because I admire our ability to weed out bullshit. Our ability to weed out bullshit is easiest in fields like mathematics where proofs are right or wrong and the unyielding judge of logic will decide the fate of your theorem or conjecture. It’s a little harder, but still relatively easy, to weed out bullshit in physics where physicists propose laws, those laws carry mathematical consequences, and the mathematical consequences of those laws are empirically tested. Once a physical law becomes widely accepted, its mathematical consequences become necessary conditions for any higher model. If you show me a glorious model that you spent 20 years analyzing and its magnetic field is not divergence free, I hate to be a jerk, but your model is bullshit and you wasted 20 years of your life.
In biology, we don’t have laws. The closest thing we have to laws are little more than empirical observations. For example, the genetic code seems to be the only genetic code used in nature, but researchers have modified the genetic code in the lab, so the genetic code is not a universal law but maybe a happy empirical accident. We have evolution, a theory that helps us contextualize how reproduction, heritable variation, and differential fitness can lead to changes in the traits of populations of organisms. Finches reproduce, so if they have different beak sizes, different beak sizes are heritable, and bigger beaks help birds get more food & reproduce more, then we trivially expect the average beak sizes in the finch population to go up. Yet, even evolution is not a law as evolution alone doesn’t have mathematical consequences help us predict much except the trivial increase of alleles that are increasing faster, limiting our ability to call bullshit. Everything is indeed ‘possible’ with evolution, so perhaps saying something is possible in nature is not scientifically meaningful.
With modern genome sequencing, however, we have extraordinary datasets of natural observations that allow us to say some things are probable and others are not. We have genomes of organisms all over the world sampled at different times and in different conditions. Our collective sampling of nature gives us a view of how evolution happens, how some mutations in some settings confer fitness advantages and become fixed in the population, and how other mutations in other settings are neutral, enabling genetic drift. We observe recombination events, from salivary amylase in humans to an Ebola gene (VP35) shared between filoviruses and bats and more. It may seem like anything is possible, but clearly things are not equally probable. Biologists’ conflation of possibility and probability stems from a lack of serious effort to scaffold the space of evolutionary possibilities into a more scientifically actionable set of probabilities. Natural origins biologists like to say something they like is possible and something they don’t like is “improbable” without actually pulling out a pencil and estimating the probabilities.
A hippo, however, probably did not hatch from a chicken egg. I haven’t checked, but I’d guess there are hundreds of millions (probably more) mutations separating chickens and hippos, and the number of mutations between parents and offspring is random but small. In theory, there’s always some non-zero probability of N mutations for any N, but at some point, like N=1 million, we should call bullshit and explore other, more likely possibilities for the origin of the hippo.
The mouse’s jellyfish gene requires more careful thinking, since technically we don’t see that exact genetic sequence in jellyfish. Weird. Maybe those N=730 mutations inserting the green fluorescent protein in a mouse were a chance event. There are so many mice! Anything is possible when you have so many mice! Of course, this path of reasoning is also littered with piles of bullshit we need to carefully avoid - a vanishingly small probability can still be vanishingly small with many mice, especially compared to alternative proximities like those derived from the grant from a lab in that same city proposing to insert these 730bp in a mouse using popular technology. If this mouse appeared 20,000 years ago, long before we knew about jellyfish genes, I’d assume it’s natural (and be very perplexed!), but when this mouse appears 1 year after the grant to make a mouse like this is proposed, it’s clear the mouse was made in a lab because there exists an alternate possibility with astronomically higher probability. It’s also informative that we saw no glowing green mice like this until immediately after the technology was developed and the grant was proposed.
We’ve lived at least 100 years without an uncontained coronavirus pandemic creating an endemic coronavirus. People have been eating guano and bats and raccoon dogs this entire time, and there has not been a newly introduce SARS coronavirus pandemic - this is in stark contrast to, say, influenza, where new strains are introduced frequently through the well-known mechanisms of antigenic shift and spillover from poultry or livestock. Suddenly, just over 1 year after a grant proposed to insert a human-specific furin cleavage site in a bat SARS coronavirus infectious clone in Wuhan, SARS-CoV-2 emerged with a human-specific furin cleavage site and restriction map of an infectious clone in Wuhan. The furin cleavage site is exactly in the S1/S2 junction where the grant proposed to insert it.
The SARS COV evolutionary tree spans over 1,000 years and there is not a single furin cleavage site to be found outside SARS-CoV-2. SARS-CoV-1 was “serially passaged” in human transmission chains during the 2002 pandemic, and yet even the later variants did not have a furin cleavage site despite humans hosting many billions of virions. The SARS-CoV-1 pandemic lasted 8 months, so with ~8 days between subsequent infections we’d estimate ~30 “serial passages” in over 8,000 patients and not a single furin cleavage site to be found. Suddenly, SARS-CoV-2 shows up with that furin cleavage site less than 1 year after it was proposed, in the exact city where virologists had the means, motive, and opportunity to do it.
Biologists need to get more serious about science or our entire field will sound like bullshit. With our theory of evolution and our massive databases of genomes, we have no excuses for abusing evolution by saying anything-is-possible and avoiding the rigor of probabilities, even those imperfectly but transparently estimated. Mutational events - whether single nucleotide mutations, insertions, deletions or recombination events - occur at well-defined rates with well-defined probabilities of occurrence per unit time, and estimating rates can help us call bullshit. If we estimate recombination events like this happen once ever 100 years, and natural origin theory claims 10 recombination events happened in SARS-CoV-2 in 10 years, then folk without the evolutionary rates could say “that’s possible” whereas folk with evolutionary rates would say “that is bullshit.” What a difference it makes to ground our reasoning in numbers instead of the fantastically infinite space of natural possibilities!
If you’re ever in doubt about whether a biological claim is bullshit, ask the biologist to show the data and their mathematical and statistical work. If there is no quantitative basis for their statement, no clear connection between a dataset and statistical methods to estimate the quantity of interest, then they are postulating and scenario-building about phenomena without any quantitative accountability, i.e. they are bullshitting. Even when we’re discussing an experimental finding - like saying “removing an FCS reduces viral titers of SARS-CoV-2 in humanized mice” - that should come with numbers. How much did it reduce the viral titers (answer: 100-10,000x)? How many mice were tested? What is the distribution of viral titers for the wild type strain? What are the odds of that happening by chance?
A sad truth, but also an opportunity, is that most biologists do not have strong quantitative backgrounds. I’m one of the unusual ones - I have undergraduate degrees in both applied mathematics and biology, and I was the first person to graduate from the first class of Princeton’s novel program in “Quantitative and Computational Biology”. It’s a tragically lonely field knowing both biology and math, yet from this vantage point one realizes numbers are the essence of hard science. Numbers are the language we use to communicate findings to compare our measurements across the world, and without which science ceases to exist. If someone doesn’t have the numbers to back up their claims, then we can’t reproduce their work or falsify their theory. Without numbers, it’s all anecdotes and bullshit, smoke and mirrors, naturalistic credulity and snake-oil making you believe anything is possible, and we can’t do science. Biologists must get more quantitative. I suspect that when they do, when the old guard of virologists who grew up in a pre-quantitative era retires and a new generation raised with computers and numbers examines the evidence anew without reputational risk, the lab origin of SARS-CoV-2 will be as clear as the lab origin of a green fluorescent mouse.
When is Nature Unnatural?
I'm always awed by your insights, intellect and POV. Thanks for another great essay. May you become rich and powerful and wipe the biology bullshitters off the face of the sad science it seems to have become.
On Point…!!
“…but still relatively easy, to weed out bullshit in physics…”
Maybe…
The Trouble with Physics-Lee Smolin
The Higgs Fake -Alexander Unzicker
Farewell to Reality-Jim Baggott
Science or Fiction-ofer Comay
Tired Light-Lyndon Ashmore
Fields of Colour-Rodney Brooks
Bankrupting Physics-Alexander Unzicker