Thread by Michael Lin, PhD-MD 🧬
- Tweet
- Mar 3, 2023
- #NaturalScience #Pandemic
Thread
Scientists can get gaslit too. If you want to see an example, read the below from a PNAS essay trying to convince us that a lab researcher picking up SARSCoV2 while collecting bat samples isn't a lab leak scenario. 
Blue-highlighted text: sliding in a flimsy argument as proof via reference, without saying what that is. If it were so logical, just present it here.
What it was: that a 2-nt difference seen in human samples from the start "proved" the virus evolved in a non-human animal.
What it was: that a 2-nt difference seen in human samples from the start "proved" the virus evolved in a non-human animal.
Red text: By shifting the boundaries of the common understanding of lab leak scenarios to exclude natural unmodified virus and include the idiotic "oh do you mean a scientist got it from non-bat wildlife", this is a straw-man fallacy trying to paint all lab leak as unreasonable
Note DOE and FBI favor lab leak of a *natural* virus. So if we consider DOE and FBI’s hypothesis a lab leak one, as seems to be the case in public understanding and reporting, then our definition of lab leak is clearly different from this author’s.
If manipulative fallacies are used, it's a bad-faith argument, not a scientific one.
On any topic, a good scientist's approach is to assign degrees of confidence to each viable hypothesis, not to prematurely reach a conclusion then try to gaslight those who don't accept it.
On any topic, a good scientist's approach is to assign degrees of confidence to each viable hypothesis, not to prematurely reach a conclusion then try to gaslight those who don't accept it.
For the record, it's okay to have initially assumed wildlife passage because of the early clusters at the Huanan Seafood Market. The FCS could be because evolution does wacky things (that's how species evolve). However lab leak was always a possibility too
3 things over time made wildlife passage relatively less likely:
1. the NEJM paper showed case #1, and 3 of the next 4 cases, had no connection to market
2. we still haven't found a wildlife host
3. WIV and WCDC still won't say if they've seriously retested samples for SARSCoV2
1. the NEJM paper showed case #1, and 3 of the next 4 cases, had no connection to market
2. we still haven't found a wildlife host
3. WIV and WCDC still won't say if they've seriously retested samples for SARSCoV2
So now we're at the point where we can't know with the available information. There are precursors and they're either in animals who never got it again or they are in lab collections which never got described again. We might never know.
Also something about this essay, phrasing the problem as two "arguments" that are fighting for "strength", and mixing scientific papers and conspiracy theory discussions, rubs me as rather unprofessional for an essay aimed at scientists.
Certain people insist I will be convinced SARSCoV2 originated from wildlife if I just listened to Worobey's talk. I have and am not impressed.
www.youtube.com/watch?v=jTJG4GFYSbQ&t=1s
www.youtube.com/watch?v=jTJG4GFYSbQ&t=1s
1. All his emphasis on cases working at market or living closer to the market (39:30) is consistent with the market being a superspreader relay location and says nothing about whether wildlife brought it there or humans brought it there (like Singapore's hawker's markets) 
2. He spends 4 min, an eternity in a talk, trying to convince us that cases being near the market proves they came from the market, but neglects to reveal that WCDC was a short walk away. This is a deliberate obfuscation because he does point out with a slide how WIV is far away
...and because he does know about the WCDC... he mentions it later as if it's not important!
Anyway he puts up this slide as if it's profound, but it's just a correlation to his chosen target. Replace "Huanan market" with "WCDC" and the answer would also be yes.
Anyway he puts up this slide as if it's profound, but it's just a correlation to his chosen target. Replace "Huanan market" with "WCDC" and the answer would also be yes.
I'm indeed wondering about WCDC more now because it's apparently what concerned DOE. (I'm curious what FBI thinks). Plus WCDC had a literally and figuratively batty video which shows an Indiana Jones mentality toward a serious topic (below).
www.youtube.com/watch?v=ovnUyTRMERI&t=2s
www.youtube.com/watch?v=ovnUyTRMERI&t=2s
3. Indeed Worobey seems to think "Lineage A" and "Lineage B" are some profoundly different lineages but the tree he shows confirms they are different by only 2 bases. The "A" root has 2 fewer mutations than the "B" root. He doesn't point this out. Yet he must know. 
And indeed he goes straight from seeing 2 lineages to concluding there were 2 jumps from wildlife days apart. That :
(a) assumes the jumps were from wildlife, not people.
(b) hides the fact that >2 mutations are being made in each host anyway, so separate events not needed
(a) assumes the jumps were from wildlife, not people.
(b) hides the fact that >2 mutations are being made in each host anyway, so separate events not needed
(c) ignores the possibility that hosts doing the spreading are human animals rather than a wildlife animals in the market. You can replace every time he says "animal" with "humans" and it would be equally valid
Essentially the entire presentation is "begging the question": assuming the answer is correct and fitting everything to that. There is no attempt to differentiate against his hypothesis and reasonable alternative possibilities of people bringing to the market.
If there is anything revelatory in this talk, it is the opposite of what his defenders thought. I learned:
A. There's indeed no ability of the data to discern case 1 being from WCDC (or elsewhere in Wuhan, still possible) vs from Huanan Market
A. There's indeed no ability of the data to discern case 1 being from WCDC (or elsewhere in Wuhan, still possible) vs from Huanan Market
B. There's no ability of the data to discern whether case 1 is in a non-human vs a human animal
C. The lineages really are completely meaningless. Intrahost variation is bigger than 2nt, so just need 1 host to create the branch.
C. The lineages really are completely meaningless. Intrahost variation is bigger than 2nt, so just need 1 host to create the branch.
Essentially the entire presentation is "begging the question": assuming the answer is correct and fitting everything to that. The best he can say is the data are consistent with a market origin. Yes I agree! But it doesn't rule out other origins
Worobey seems to be (nervously) hoping that we just don't think of other plausible hypotheses. By talking for 45 minutes and concluding a market origin before even mentioning the colocalized WCDC, it would seem he knows he is overstating his case, and just hopes we don't notice.
Backing up the intuition that you can't use statistical analysis of a sparse dataset to claim proof that the observed distribution came from passage through wildlife. Sometimes if a claim is hard to understand it's because it doesn't make sense.
arxiv.org/abs/2208.10106
arxiv.org/abs/2208.10106
Funny that a few vocal responders insist that not accepting Worobey's claims as definitive proof is evidence of my being unqualified to judge.
Besides the intellectual homogenization issues, there's the question if they'd impose the same judgement on Fauci who keeps an open mind
Besides the intellectual homogenization issues, there's the question if they'd impose the same judgement on Fauci who keeps an open mind
And this is not an appeal to authority by me. As long-time followers know I've always had a neutral stance that we l don't have the data to rule out either commercial or research activity as the conduit to Wuhan. That's been my position on logic+data alone, before Fauci said so
So if someone denigrates a scientist of a certain rank or stature for disagreeing, but doesn't do the same for another scientist of a higher position, then it's no longer a scientific discussion but a political one
As for publications, I think it's valuable to have the models published so we can discuss them, and to gather all available data in one place. But given the serious rebuttals by statisticians, our ability to say we know the history of early nonsequenced cases from them is limited
