3 ms·
> We did not find the 19-nucleotide sequence CTCCTCGGCGGGCACGTAG in any eukaryotic or viral genomes except SARS-CoV-2 with 100% coverage and identity in the BLA
by flobosg 5y ago
> We did not find the 19-nucleotide sequence CTCCTCGGCGGGCACGTAG in any eukaryotic or viral genomes except SARS-CoV-2 with 100% coverage and identity in the BLAST database (Supplementary Tables 1–3).
Huh? I just run BLAST (blastn, specifically) with that nucleotide sequence and found several eukaryotic genomes containing it in non MutS-like contexts, aligning sense and anti-sense strands, with 100% coverage and 100% identity. Besides, there is no “BLAST” database (I used the non-redundant -nr- one, for instance; BLAST is just the name of the tool). One possible reason they didn’t find the sequences is that BLAST, by default, only returns the first 100 hits, which in their case might have been prokaryotic ones. I didn’t check viral genomes but might later.
Disclaimer: I know and have used BLAST since my undergrad years, about a decade and a half ago.
- y4mi 5y agoThat sounds more like an issue in the reporting being unclear about the constraints used for the identification then the actual finding.
- flobosg 5y agoExactly, and that could change the interpretation of their findings completely. I haven’t checked their credentials, but I would be surprised if any of the authors has a background in bioinformatics or computational biology. BLAST is used on a daily basis by a lot of scientists (its paper is among the most cited of all time), but many of them lack the theoretical knowledge to analyze in depth their results. This paper doesn’t have a single line about methodological details, such as the parameters used in their BLAST runs. The authors don’t even refer to P- or E-values either, despite being prominently displayed on the output table of every run!