BA.2 sublineage with S:I68T and C7471T circulating in UK (568 seqs globally)

[FedeGueli]

Follow up from #524
S:I68T was spotted by @olias120676 in several european clades
Here i want to propose the one carrying S:I68T and then C7471T
It reached 376 sequences mainly in England and Wales.
https://cov-spectrum.org/explore/World/AllSamples/Past6M/variants?aaMutations=S%3AI68T&nucMutations=C7471T&pangoLineage=BA.2*&aaMutations1=S%3AI68T&nucMutations1=C7471T&

Usher tree is clear:

Growth advantage over BA.2 is moderate around 22% in UK

[c19850727]

Hi @FedeGueli , here's a cluster of potential BA.1*+BA.2* recombinants (5 seqs, with breakpoint at NSP3) that interestingly have a conserved Spike I68T substitution ((and another unique N:Q7K substitution):

EPI_ISL_11384598, 11679909, 11814006, 11938911, 12070160

https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_10615_bde6b0.json?c=userOrOld&label=nuc%20mutations:T21765C,C28292A

[corneliusroemer]

Thanks for the proposal.

If S:68T is potentially homoplasic. So Usher can't necessarily be trusted on the sequence of mutations. It's hard to tell whether C7471T came first or second. Both is possible.

Albeit I slightly favour C7471T first because that mutation was present in the UK before.

[corneliusroemer]

This keeps growing in the UK, I'd be in favour of a designation of C7471T + S:I68T.

[FedeGueli]

List of sequences:
contributors (5).csv

This sublineage goes on growing: 507seqs today
@chrisruis @corneliusroemer @AngieHinrichs

[corneliusroemer]

This is about a quarter of the UK BA.2 + S:68T. One needs to decided whether there'll be one big BA.2 + S:68T lineage, or whether we designate many BA.2 + S:68T + defining-distinction.

This Belgian #565 makes up about half of all global S:68T, the rest remains undesignated. I've proposed two other regionally distinctive ones in #609 #610

[FedeGueli]

568 seqs as today

[AngieHinrichs]

FWIW in the current UCSC/UShER tree this is a branch of BA.2.36 (BA.2 + C22792T + T21765C/S:I68T) although as Cornelius says either C7471T or T21765C/S:I68T could have come first.

But there's plenty of evidence for C7471T having homoplasies of its own. There's a branch with 282 sequences with C7471T and no change on S:I68 (BA.2 + C25416T > G26577C > C7471T, starting with MMR/DSMRC-095/2022|ON038768.1|2022-02-19) but those are mostly from Myanmar/Thailand/Japan/Australia. There are a dozen French sequences (e.g. France/ARA-HCL022054062502/2022) with BA.2 T9866C > G29332T > C7471T, 15 sequences from Denmark (e.g. Denmark/DCGC-417690/2022) with BA.2 + C22792T > C25624T > C14925T > C1594T,C12043T > C7471T. And there are 86 mostly USA sequences (e.g. USA/NJ-DHSS-B1181997/2022) with BA.2 + C22792T > C25624T > A24433G > C22450T > C7471A,A18010G. (BA.2 + C22792T > C25624T = BA.2.9) To find sequences from England with 7471T and not S:I68T I had to do a cov-spectrum query, and they're scattered all over BA.2 (in branches with fewer than 10 descendants which is the threshold I used in taxonium's mutation search), with lots of other mutations.

So no harm in calling this a sublineage of BA.2.36, I think.

[FedeGueli]

Thank you @AngieHinrichs !
i'll close this one after your clear explanation as one syn mutation isnt enough to designate a sublineage.