So, Flo Debarre et al’s raccoon dog analysis that has caused such a media frenzy has been released and what does it show ?
Not much 🧵
2/ Essentially, it confirms Gao et al’s preprint analysis that there was nuc acid from animals in addition to humans in the samples 👇 (no surprise there)
It adds some species specific info
3/ The analysis is crude, and taxonomic attribution method naïve
They rely on seq assembly, which miss a lot of info
They use numbers of assembled contigs as a metric for quantity of species specific material
This is semi-quantitative at best (due to the vagaries of assembly)
4/ Even then, they fail to see if the species specific material correlates with numbers of SARS2 reads, which is inexplicable
5/ They apparently identify a potentially new subspecies of raccoon dog – this would worth following up for forensic purposes, but bafflingly they fail to do so
6/ But they also seem to identify new subspecies of masked palm civet, hoary bamboo rat, Malayan porcupine and Amur hedgehog. This could be interesting, or more likely indicates problems with their assembly. Oddly, they fail to follow this up and validate their assembly qualities
7/ Bizarrely, they seem unable to differentiate between DNA and RNA (hint, trying mapping reads to annotated mito genome or nuc genome). This has importance as it can affect estimates of relative species proportions
8/ Tellingly, none of the stalls with raccoon dog nuc acid have a human SARS2 case linked to it (which puzzlingly they fail to mention)
(figure on left from our HSM Zoonosis critique, on right from Debarre et al)
9/ This saga is a case study in the perils of making grandiose claims without having completed the analysis, and of the hubris to embark into a new subject area without specialists (metagenomic) to scrutinize and suggest robust analyses
2/ @quay_dr uploaded our original preprint on March 26
We identified 7 'frozen' SARS2 sequences from UNC, that were basal to other sequences generated at the same time. These could represent early SARS2 strains that had escaped from a research facility
3/ Before uploading the preprint, we sent an email to Dirk Dittmer and Melissa Miller of the Clinical Molecular Microbiology Laboratory, UNC Hospital, regarding the 7 anomalous sequences we had detected from their sequencing facility
Steve Quay @quay_dr and myself have published evidence for a potential lab leak of SARS2 at the University of North Carolina (UNC)
7 genome sequences collected by UNC in 2020-21 are basal (ie 'frozen') indicating a potential lab origin 🧵 zenodo.org/records/150833…
2/ The 7 sequences were collected from Jun 2020 to Jan 2021 and submitted by Dr Dirk Dittmer of the Clinical Molecular Microbiology Laboratory, UNC Hospital
Puzzlingly, they show strong similarities to the SARS2 reference sequence Hu-1, sampled in Dec 2019
3/ On the dates the 7 sequences were collected, they should have accumulated more SNVs than they actually possess, according to the SARS2 molecular clock
I've published a paper on the emerging phenomenon of 'frozen' virus genome sequences, and how they can indicate lab leaks
"Frozen' virus genome sequences are generated from recent outbreaks, but show surprising similarity to historic strains 🧵 mdpi.com/2076-2607/12/1…
2/ The classic example is 1977 'Russian' H1N1 Flu, which re-emerged after having gone extinct in the 50s. Genomic sequences were identical to strains from the 50s, indicating it had been stored for > 20 years before release nature.com/articles/27433…
3/ An alternative scenario of viral host persistence / latency seems unlikely
A clue to its non-natural origin was that only younger people got sick, while older people were immune. This is because the latter had been exposed to H1N1 in the 50s and before pmc.ncbi.nlm.nih.gov/articles/PMC23…
It has been reported German Intel has assessed that COVID19 originated via a lab leak
Swiss paper NZZ reports that our discovery of a MERS-related infectious clone with a chimeric spike from Wuhan sequencing datasets was considered by German Intel 🧵
2/ The study by @humblesci @Daoyu15 @BiophysicsFL @ydeigin @quay_dr and myself was published last year and demonstrated unreported GOF experimentation on novel MERS-related coronaviruses in Wuhan prior to the pandemic fortunejournals.com/articles/disco…
3/ Concerningly, we found evidence that a MERS spike (with FCS) was inserted into a HKU4r-CoV backbone, likely enhancing transmissibility in human cells, as the RBD is expected to have a high binding affinity for the hDPP4 receptor
Here is a thread on the preprint version of the paper:
Pekar et al have published a paper claiming that 2 SARS2 A/B intermediates from Feb 2020 are derived rather than ancestral
The existence of ancestral intermediates would falsify Pekar et al's earlier claim of 2 independent spillovers of lineages A and B at the Huanan Market 🧵
2/ Hence Pekar et al undergo extraordinary convolutions in order to try and claim the 2 A/B intermediates, identified by Lv et al 2024, are late arrivals rather than the original intermediates that led to the evolution of lineage B from lineage A
3/ To sum, they once again (ab)use modeling to support intuitively unlikely scenarios
They claim it is more likely that the 2 Lv et al intermediates arose from either Hu-1 (lin B), or its closest lin A relative (C8782T,T28144C) AFTER the 2 lineages had originally split
2/ This striking conincidence indicates that the novel pangolin-related coronavirus we identified from contaminated datasets (GX_ZX45r-CoV) was likely part of an infectious clone (IC), given that one of the reads was ligated to pUC57
From our paper:
3/ This can be inferred given that ZLS et al report using pUC57 for cloning genome fragments of HKU5 coronavirus prior to assembly into an IC in their recent Nature paper on engineering coronavirus receptors