BL.1 sublineage with S:490L, Switzerland, Canada, Germany [7 seqs]

[Sinickle]

Description
Sub-lineage of: BL.1
Earliest sequence: 2022-8-25, Switzerland — EPI_ISL_15036254
Most recent sequence: 2022-9-9, Canada — EPI_ISL_15054432
Countries circulating: Switzerland, 4. Canada, 1.
Number of Sequences: 5
GISAID Query: Spike_R346T, Spike_F490L, Spike_D574V
CovSpectrum Query: Nextcladepangolineage:BA.2.75* & S:346T & S:574v & S:490L
Substitutions on top of BL.1
Spike: F490L
ORF1b: P2612L, R2613N
Nucleotide: C21302T,G21305A,C21304A,T29867A,T23030C

USHER Tree
https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_daf3_1bd4e0.json?label=nuc%20mutations:G22599C

Evidence
We already have designated a couple sublineages of BA.2.75 with an S:490S mutation (BN.1, BM.1.1). Although S:490S is the most common S:490 mutation in BA.2.75*, it appears that the site is under selection pressure in general, and there are also some sequences with S:490V and S:490L.
This will possibly add additional advantage to BL.1, which is a fairly fit and large lineage already.
I'd like to call attention to the peculiar ORF1b mutations, at position 2612,2613. These appear to have happened prior to the S:490L mutation. I haven't looked into this too closely, but it seems this pair of mutations has happened on many Omicron lineages, and more than 3000 sequences in the last 6 months.

Genomes:EPI_ISL_15036254, EPI_ISL_15036087, EPI_ISL_15036334, EPI_ISL_15063485, EPI_ISL_15054432

[FedeGueli]

One more sequence from Germany uploaded today @Sinickle . Tot is 6 now

[shay671]

So this has been designated? @corneliusroemer is it accepted?

[jknappen]

So this has been designated? @corneliusroemer is it accepted?

Yes, it is BL.1.3 (click on the hex number and than open the first diff to see it)

[ryhisner]

@Sinickle, the ORF1b:R2613N mutation creates the sgmRNA AAACGAAC pattern noted by @theosanderson and @thomaspeacock in their paper "Emergence of new subgenomic mRNAs in SARS-CoV-2." The ORF1b:P2612L I believe increases the homology of the nucleotides on either side of the AAACGAAC. I frankly don't understand how this works except that somehow, if the nucleotides on each side mirror each other, the sgmRNA works better.
https://www.biorxiv.org/content/10.1101/2022.04.20.488895v1

[shay671]

Hi @ryhisner , Those 3 nucsubs appear in a lot of variants

and also specifically in BA.2.75

It might be that this is Ultra converging mutations go all together each time. Or this is some technical problem.

@AngieHinrichs, what do you think ?

[AngieHinrichs]

I think in most cases, three (sometimes more) substitutions adjacent or nearly adjacent to each other are errors -- although B.1.1's 3-nucleotide substitution of 28881-28883 is very real, so one can't dismiss them all without a proper look. I see them pretty regularly when investigating lineage assignments that don't look right. I call them "run-of-nucleotides" cases, not sure if there's a better term.

If a run-of-nucleotides branch also contains multiple other very familiar mutations that define other sublineages (or at least, are seen in sibling branches), then I think that is pretty convincing evidence that the run-of-nucleotides is a sequencing error. So let's take a look at those in BA.2.75.

The 2022-10-10 UShER tree has a branch BA.2.75.3 > C21302T,C21304A,G21305A, but most of the sequences on that branch have a reversion on BA.2.75.3-defining 18583, so back to BA.2.75. Then there's a branch with BA.2.75.1-defining A23283T (S:D574V). And within that branch is the entire sublineage BL.1.3, which should be a sublineage of BA.2.75.1 > BL.1 not BA.2.75.3.

That looks clearly incorrect to me -- those sequences past the 18583 reversion don't belong in BA.2.75.3, and the sequences past the A23283T belong in BA.2.75.1. But in parsimony-land, the attraction of 3 shared substitutions (21302,21304,21305) is very strong.

I have recently added a lot of branch-specific masking to BA.2.75, mostly to get rid of reversions on BA.2-defining and BA.2.75-defining sites. I think I'll mask those as well. They crop up a fair amount in BA.5 as well (actually over the entire tree except Delta... ah, because they are already masked in Delta, by Cornelius's request after they caused artefactual retracted lineage AY.89, see #398).

Perhaps I should mask 21302, 21304 and 21305 in the entire tree. Any objections?