Thread by @jbloom_lab

I am getting lots of questions if my pre-print about some #SARSCoV2 sequences that were removed from Sequence Read Archive tell us anything about lab accident versus natural zoonosis.

I posted summary of pre-print below, but did not directly address this point explicitly (1/n)

The answer is NO. The people using it to strongly support either argument are those that have become so emotionally invested in their opinion that they have lost the ability to analyze anything objectively outside of the framework of that argument. (2/n)

What the pre-print does imply is as follows:

First, there may be additional relevant data in obscure locations that aren't the places where we are accustomed to looking (e.g., on the Google Cloud, in table 1 of a paper on diagnostics, etc): (3/n)

Second, in my opinion, anybody doing phylogenetics on early #SARSCoV2 sequences from China should spend as much time on metadata as algorithms. Sequences in databases may be non-representative. Sequences collected in Guangdong might be from infections from Wuhan. Etc. (4/n)

Third, preprint provides modestly more evidence for progenitor being in clade A (not market clade), & substantially more evidence it might have T at site 29095. However, current inferences are likely based on incomplete data. @sergeilkp says it best: (5/n)