Still undesignated Orf1a:K261N Delta sublineage is dominant in Russia and widespread in Europe

[FedeGueli]

Orf1a:K261N spotted in various issues here ( Ex.: #268 )
is the dominant lineage in Russia according to COVSpectrum:

In Russia the share of "non AY.s" B.1.617.2 is 47,60% in the last three months and 82,1% in the last week.

Of these 733 sequences 90,18% has Orf1a:K261N

https://cov-spectrum.ethz.ch/explore/Russia/AllSamples/Past3M/variants?pangoLineage=B.1.617.2

And it is widely spread in several european countries:

Country Total Variant Sequences
Belgium 988
Bulgaria 126
Canada 253
Czechia 506
Denmark 5234
Estonia 857
France 1834
Germany 4506
Iceland 44
Ireland 303
Israel 255
Italy 998
Japan 51
Lithuania 175
Luxembourg 193
Montenegro 26
Netherlands 1219
Norway 776
Poland 851
Russia 661
Slovakia 283
Sweden 3598
Switzerland 375
United Kingdom 2410
United States 6403
Croatia 166
Ecuador 75
Georgia 48
Greece 291
Guadeloupe 9
India 34
Spain 153
Turkey 571
Slovenia 439
South Korea 160
and so on.

I think it is very urgent to designate it to understand well the actual wave in Europe for citizens and scientists.

And keep in mind that ~14% of AY.43 sequences in europe are missassigned carrying instead the Orf1a:K261N mutation distinctive of this lineage
https://cov-spectrum.ethz.ch/explore/Europe/AllSamples/Past3M/variants?pangoLineage=AY.43

@chrisruis @AngieHinrichs @corneliusroemer

[FedeGueli]

It is a self evident clade.

[MCB6]

Of these 733 sequences 90,18% has Orf1a:K261N

we discussed it at length in the Orf7a:P45L thread, only I used the GISAID's standard NSP2:K81N designation for it. Basically Orf1a:K261N is called in almost all Russian samples, but (older) Orf7a:P45L - only in some of them. But it isn't a reversion mutation back to Proline, as @corneliusroemer showed. Rather, it is a possible genotyping problem combined with a flaky phylogenetic tree building, so some the supposedly Orf7a:45P genomes are scattered across many branches.

I am just using NSP2_K81N to query Russian uploads on GISAID now. Out of most recently collected 100 samples, 20% are classified as AY.43 and most of the rest as B.1.617.2.

My gut feeling is that epidemiologists in Eastern Europe don't realize that there is a strong strain continuity since spring or summer, because Pango labels keep on shifting, and that it may help them if they realize that they are dealing with the same strain for a second wave in a row, rather than with a changed variety of the virus.

[MCB6]

PS: adding graphs. Orf1a:K261N aka NSP2:K81N has been dominant in Russia since June, flat at 10-25% rate in a number of East and Central European countries, but interestingly, it's in retreat in Finland even as they experience a fall wave


https://cov-spectrum.ethz.ch/explore/Russia/AllSamples/from=2021-06-09&to=2021-11-14/variants?aaMutations=Orf1a%3AK261N&pangoLineage=B.1.617.*

[FedeGueli]

Thx @MCB6 for your analysis.

There are a lot of sublineages of K261N , but it is becoming hard to follow them without a designation.
Plus missassignements create rumour and we dont need that, Europe is hardly hit right now and i dont see how to keep this clade undesignated.

Missassignments of it arent just in AY.43 so it creates an unnecessary mess:

[MCB6]

There are a lot of sublineages of K261N , but it is becoming hard to follow them without a designation.

I checked where else it reached high %age levels, and what was the relation with Orf7a:P45L. Most of the countries were Russia's neighbors - Finland, Estonia, Latvia, Ukraine - but also Israel. The highest fractions of Orf1a:K261N were recorded during the lull before the summer waves, beginning to grow very fast in June. One hypothesis was that it might have been related to the travels of the soccer fans of UEFA Euro, but the dates of the spread appear to be a week or two too early to fit this idea.
Orf7a:P45L is common among the K261N sequences in all of these location, with the notable exception of Israel, where Orf7a:P45L share quickly dropped to low levels (2nd image). A reasonable phylogenetic analysis may help with figuring out what happened, but since @corneliusroemer warned about issues with some phylogeny tools, I didn't even attempt playing with it, leaving it to the people who know the intricacies of phylogeny tools better.

[MCB6]

Since Israel experienced such a wild swing from mostly Orf7a:P45L to mostly negative genotypes, I decided to check USHER trees for various timepoints of the Israeli Delta wave. I wondered if I will clearly see two distinct branches.
The June collection dates yield the following (one minor but apparently separate branch is visible)

Restricting to Orf7a:P45L only, I see only the "lower sub-branch" of the main branch

And looking at the October genomes, I see the same branches as in June, just shifted to the right a little, of course

So it looks like most samples with or without Orf7a:P45L are closely related (although there may exist a minor completely unlinked branch without it). Formally these two types of samples mostly map to adjacent sub-branches, but I can't tell if this separation is a tree-building artifact. But at the very least they look closely related.

[chrisruis]

Thanks @FedeGueli We've added this as AY.122 in v1.2.95 to start on the branch with G1048T (Orf1ab:K261N). From what I can see, it looks like Orf7a:P45L drops in and out across the clade so this doesn't look like an ideal marker for a sublineage so for now we've designated the complete clade as AY.122 with the exception of the clade that was AY.12.

Lineage AY.12 clusters within the AY.122 clade. We've therefore renamed AY.12 to AY.122.1 and withdrawn the AY.12 label. AY.122.1 is downstream of C23934T (S:T791I).

[FedeGueli]

Thank you vey much @chrisruis , good choice on AY.12 'renaming'

Just to say all the credits about this lineage go to @c19850727 🙏

There are a few sublimeages already proposed for AY.122.

i'll share the issues further below so you could monitor them.easily.