The whole likelihood basically hinges on the fact that the outbreak occurred in Wuhan and that the Wuhan Institute of Virology has been working for decades on enhancing coronavirus strains. That's quite strongly circumstantial but it's not evidence. Possible chimerization and furin-cleavage insertion seem a lot more interesting imo but are weighted much lower.
Based on their report [1], most of the likelihood of lab-escape (almost 50x weight) just stems from the fact that the outbreak is in Wuhan. They state that it's because of the proximity to Wuhan Institute of Virology (WIV) and the lab's gain-of-function research - only one of 5 locations world-wide.
That single "bullet-point" re-weights zoonotic origin from 97% to 56% and lab-escape from 1.4% to 42%. Otherwise their final likelihoods would be: "zoonotic" 85.5%, "lab-escape" 8.5%, "bioweapon" 6%.
[1] https://www.rootclaim.com/analysis/what-is-the-source-of-cov...
>Furin cleavage sites are not common in other related coronaviruses.
However, this claim appears to have been investigated and debunked [1]
>Furin cleavage sites occurred independently for multiple times in the evolution of the coronavirus family, supporting the natural occurring hypothesis of SARS-CoV-2.
[1] https://www.sciencedirect.com/science/article/pii/S187350612...
The initial prior of zoonotic origin simply because that was usually the case in the past is just as circumstantial, but also just as solid in bayesian terms.
The Harbin Veterinary Research Institute in Heilongjiang is another BSL-4 lab
The posting says the startup "claims COVID-19 originated in a lab", whereas the Twitter post says it "has likely originated in a lab."
There's an important distinction between fact and likelihood of fact.
No idea how credible that analysis is, but it seems to me like it's done in good faith.
At best this is a startup company seeking publicity so 'good faith' is to be taken with more than a grain of salt.
They're stating the facts of the results of the probabilistic analysis, not the facts of the actual situation. Their only comment about the actual situation is that it's "likely".
I can see how it could be interpreted in the way you suggest, but if you read it from the perspective of discussing the probabilistic analysis, I don't think they're intending to mislead. But, again, I also don't know how strong the analysis is. It could be that the analysis is weak, in which case I still wouldn't think the tweet is likely deliberately misleading, but simply wrong.
> 11 November 2020 Editor's Note: Readers are alerted that concerns have been raised about the identity of the pangolin samples reported in this paper and their relationship to previously published pangolin samples. Appropriate editorial action will be taken once this matter is resolved.
Is it a fact that this virus has "anomalies" in its DNA?
If the answer is 'no' then there is no point discussing their 'analysis' further.