search

cov-lineages / pango-designation

Repository for suggesting new lineages that should be added to the current scheme
Other
1.04k stars 97 forks source link

Change defining mutation of BA.2 from A22786T to A22786C #410

Closed lenaschimmel closed 2 years ago

lenaschimmel commented 2 years ago

(I'm creating a new issue instead of commenting under #361 because this affects information in multiple other issues.)

In #361 the defining mutations for BA.1 and BA.2 are listed as a table. Both the image and the linked csv file contain A22786T.

I recently stumbled across many BA.2 sequences with A22786C instead of A22786T. It was confirmed by @FedeGueli by a CovSpectrum query that A22786C is much more common than A22786T (822 times more common at time of writing). I think this means that the table is wrong. All eleven samples with T occurred between 2021-12-19 and 2021-12-30 which indicates that C was also dominant when the tables were created on 2021-12-07.

I've checked the other issues and found that #367 and #392 also list A22786T as mutation of BA.2. #390 does not explicitly mention it, but the image in this comment gives further proof that A22786C is the actual mutation.

In this repository some files inside misc/B.1.1.529_sublineages are affected and must be updated, also some files in the gromstole repository by PoonLab / @ArtPoon. Where else could this have consequences?

Luckily both potential nucleotide mutations result in S:R408S. (So at least I'm not worried that Biontech is currently producing BA.2-vaccine with the wrong spike protein 😅) Do you know of any BA.2-specific PCR primers/probes that may be affected?

corneliusroemer commented 2 years ago

I double checked and it also appears to me that BA.2 has 22786C instead of 22786T.

There are just 11 BA.2 with T, all from Denmark end of December.

image https://cov-spectrum.org/explore/World/AllSamples/Past6M/variants?nucMutations=22786T&pangoLineage=BA.2*

ArtPoon commented 2 years ago

Thanks @lenaschimmel for pointing this out. Our BA.2 file was derived from the CSV linked from issue #361 of this repo. I'll log a new issue to address this on our side.

lenaschimmel commented 2 years ago

Via Google Scholar I could only find one article that is potentially affected by this: In silico evaluation of the impact of the Omicron variant on the sensitivity of RT-qPCR assays for SARS-CoV-2 detection using whole genome sequencing. There, A22786T is listed in a table that explicitly names #361 as the source. As far as I can tell by a quick look at supplementary file 3, the findings of that study should not be affected by the kind of mutation at position 22786, as only the probe of 8_HKUniv_S comes even close to it but does not overlap.

AngieHinrichs commented 2 years ago

Thanks @lenaschimmel for finding this! I think it was my dumb error when adding nucleotide mutations to @rambaut's list with protein-coding mutations, sorry about that. I corrected the file pango-designation/misc/B.1.1.529_sublineages/Omicron_BA.1_BA.2_mutations.csv but the accompanying PNG and PDF files (and the PNG in #361) will need to be updated by @rambaut.

lenaschimmel commented 2 years ago

Thanks! I think the table in this comment on #367 should also be updated by @rambaut.

And the image in #382 #392 by @bitbyte2015.

I think it would be important that the change is visible when looking at #361. The conversation for #361 has been locked and and limited to collaborators - I'm not a collaborator, so I can't comment there. (Also, me mentioning #361 here does not show up there as it normally would.)

FedeGueli commented 2 years ago

@lenaschimmel it was #392 by @bitbyte2015

lenaschimmel commented 2 years ago

I just opened to PRs to update PNGs, PDFs, CSVs and a Markdown file. If I'm not mistaken, issue #361 will then contain/link the updated files automatically, so that @rambaut does not need to edit anything.

Issue #392 still needs to be edited afterwards. Not sure if @bitbyte2015 must do it, or if other admins / maintainers could go ahead. I still would appreciate it if someone posts a comment under #361 so that the timeline of changes is obvious for anyone who reads the discussion.

rambaut commented 2 years ago

Thanks. Great help.

lenaschimmel commented 2 years ago

@rambaut You're welcome!

If I'm not mistaken, issue #361 will then contain/link the updated files automatically, so that @rambaut does not need to edit anything.

I think I was mistaken, and you still have to edit the issue manually to use the new files :/

corneliusroemer commented 2 years ago

I think this can be closed, the changes have been implemented