Sister Group of EG.5 being probably successful in Israel, 98 seqs total at 2023-06-01

[krosa1910]

Sublineage of: XBB.1.9.2
Mutations Acquired: Seems to be simply T22928C
Gisaid Inquiry: A16878T( to identify it as 1.9.2),22928C
Lineage Collections: EPI_ISL_17256318, EPI_ISL_17344757, EPI_ISL_17349171,
EPI_ISL_17418267, EPI_ISL_17422205, EPI_ISL_17469602,
EPI_ISL_17469700, EPI_ISL_17470513, EPI_ISL_17483735,
EPI_ISL_17483758, EPI_ISL_17483773, EPI_ISL_17483792,
EPI_ISL_17483821, EPI_ISL_17483838, EPI_ISL_17483857,
EPI_ISL_17483919, EPI_ISL_17483932, EPI_ISL_17483968,
EPI_ISL_17484004, EPI_ISL_17484007, EPI_ISL_17516620,

Earliest: EPI_ISL_17256318 Germany 2023-03-09
Latest: EPI_ISL_17483857 Israel 2023-04-04
Distribution: 1 Canada, 1 England, 1 France, 4 Germany, 1 Netherland, 13 Israel
As what I described in #1949 , when checking trees with S:456L, I discovered three medium trees. The last of them is a tree within xbb.1.9.2. I was trying to propose it but later I found that some of its sequences are already proposed in EG.5.
https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/singleSubtreeAuspice_genome_1b5a5_6081b0.json?f_userOrOld=uploaded%20sample （Total Structure of the tree with F456L）

Yet, what I initially naively treated as a "tree" was actually an agglomeration of two trees. L is an amino with 6 possible codon, and three such codon series are easily accessible form the triple U codon series from 22928-22930. The transition at 22928 from T to C is easier, and that's why it was chosen by FE.1,FD.1.1 and XBB.1.5.10, but EG.5 chose much rarer T22930A.
https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_3dab1_619c30.json?f_userOrOld=uploaded%20sample （EG。5 Tree）
However, a smaller tree existed below EG.5's tree, and that tree originate from the familiar T22928C path. Furthermore, there is evidence to suggest that it reach good share in Isarel that is far from Germany as where it could originate from, and could prove to be successful.
https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice2_genome_3dab1_619c30.json?f_userOrOld=uploaded%20sample （Small Tree，not sure why usher did its frequent wrong job of adding outgroup sequence）

[FedeGueli]

@krosa1910 great catch, could you add the usher tree link please to each of your proposals? it would be much easier to first check the lineage. thx

BTW thank you very much for the two routes to 456L very informative and useful for wastewater recognition or recombinants

[FedeGueli]

@krosa1910 did you find the original EG.5 issue? It is #1918 and it contains two more sublineages of XBB.1.9.2 with 456L / T22928C

[BorisUitham]

Yes, three lineages were described but only the biggest was designated. I also noticed the difference in nucleotide but only mentioned it shortly. I wonder what the ratio between the two nucleotides are across various 456l lineages

[FedeGueli]

@BorisUitham i suggest to open new issue for the other teo if you see they are growing too

[BorisUitham]

@FedeGueli I think this proposal is that already. I see the second biggest (israel+uk) and third biggest (germany, france, netherlands) are combined in this proposal, I found them as separate. I cant check right now if they are actually two different trees, but the difference was basically q146k in the german+french+dutch ones which sometimes is an artefact, right?

[krosa1910]

Yes, three lineages were described but only the biggest was designated. I also noticed the difference in nucleotide but only mentioned it shortly. I wonder what the ratio between the two nucleotides are across various 456l lineages

I quoted that transitions in general are easier than transversion, so applying that conclusion to F456L, it means that the transversion T22930A would be less likely than T22928C. I think this is also clear from the fact that 22928C did arise multiple times with decent success (early 1.5.10, later FD.1.1 and FE.1 both reaching borderline success, growth comparable with EG.1 which I consider to be a great variant under monitor) , while it seems first time that T22930A is booming with EG.5. But for the virus duplication the effect would be similar, and EG.5 and this proposal seems to have advantage of arising in the already successful XBB.1.9.2 with additional good mutations. (I have tried to conduct comparisons of the two variants under 1.9.2, both is capable of growing at two digit rates, that makes them probably even comparable with the faster lineage we currently find, like 1.16 and 2.3.2)

[krosa1910]

@FedeGueli I think this proposal is that already. I see the second biggest (israel+uk) and third biggest (germany, france, netherlands) are combined in this proposal, I found them as separate. I cant check right now if they are actually two different trees, but the difference was basically q146k in the german+french+dutch ones which sometimes is an artefact, right?

Not really sure by q146k. In your original picture it is pretty clear that 146k is a thing, but in my search they are non-existent. I doubt that it is some form of artefact then. @BorisUitham If you can provide the initial Usher link it would be more helpful.

Also, I think that anyway your proposed second tree is more interesting/subtree of my proposal. Not only because it spread out in a foreign region and It contained one of the most positively rated mutations within Orf6 from jbloom calculator. Not sure what effect it have though.

[alurqu]

Yes, three lineages were described but only the biggest was designated. I also noticed the difference in nucleotide but only mentioned it shortly. I wonder what the ratio between the two nucleotides are across various 456l lineages

I quoted that transitions in general are easier than transversion, so applying that conclusion to F456L, it means that the transversion T22930A would be less likely than T22928C. I think this is also clear from the fact that 22928C did arise multiple times with decent success (early 1.5.10, later FD.1.1 and FE.1 both reaching borderline success, growth comparable with EG.1 which I consider to be a great variant under monitor) , while it seems first time that T22930A is booming with EG.5. But for the virus duplication the effect would be similar, and EG.5 and this proposal seems to have advantage of arising in the already successful XBB.1.9.2 with additional good mutations. (I have tried to conduct comparisons of the two variants under 1.9.2, both is capable of growing at two digit rates, that makes them probably even comparable with the faster lineage we currently find, like 1.16 and 2.3.2)

The observational results of Bloom and Neher also show that, as you expected, across all clades T-to-C transitions occur at a faster rate than T-to-A transversions. See https://github.com/jbloomlab/SARS2-mut-spectrum/blob/main/results/synonymous_mut_rates/rates_by_clade.csv

[BorisUitham]

@krosa1910 i concur, there is also a new one today from japan (edit: and canada) besides uk and israel so thats quite a wide spread even if theres still so few sequences.

As for the usher tree link, im sorry i dont have it anymore. But what I basically did was gather all the xbb.1.9.2* + 456l on gisaid, and examined the downsampled global tree of that on usher, to see multiple convergent mutations within the lineage without too much clutter from other sequences

[FedeGueli]

With T22928C I see:
1 a singlet from Canada defined by orf1a:P971L
2 a 6 seqs lineage from Germany with S:Q146K
3 a 8 seqs from Israel et al with C1684T and C17766T plus Orf6:Y49H

with T22930A i see:
1 EG.5 that notably has gained N:L161F
2 one singlet from Indonesia with many mutations S:Q52H,N:G215S,Orf1a:A690V, orf1a:A3143V
with a quadruplet at the end of Orf1b tht makes me think it could be a TRS (cc @ryhisner EPI_ISL_17308775)

Here the tree:

https://nextstrain.org/fetch/genome.ucsc.edu/trash/ct/subtreeAuspice1_genome_27371_6fb270.json?c=gt-S_456

[BorisUitham]

This is the query i made earlier for the israel branch (20 seqs and growing): A27507C, T22928C, C1684T
This is the one i made for the german/french/dutch one: (6 seqs, a lot less activity)
A27507C, T22928C, C21998A

[krosa1910]

This is the query i made earlier for the israel branch (20 seqs and growing): A27507C, T22928C, C1684T This is the one i made for the german/french/dutch one: (6 seqs, a lot less activity) A27507C, T22928C, C21998A

Feel like I do not trust 146K/Q difference now, so would suggest a merge of these two. But thanks for sharing of query information

[krosa1910]

@BorisUitham @FedeGueli
Gisaid just uploaded a very old sequence:EPI_ISL_17537558 that is collected in 02-14, Canada but submitted yesterday.
I thought that this could be a typo from 04-14 but was not sure. Could you help me decide upon that.
This would mean a potential change of earliest lineage for this proposal, and even #1918 as it is earlier than the Indonesian lineage of EG.5( although we already know that the proposal is polyphyletic)

[krosa1910]

A certainly big upload happened this week。 Now for updated total count:
1 Belgium (new), 30 Canada,1 England, 1 France, 4 Germany, 13 Israel, 11 Japan, 1 Netherland, 1 Scotland, 1 Shanghai, 1 Singapore, 3 Spain, 8 US.
Canada submitted 16 seqs this week, and that is a lot compared to the 93 total seqs it submited this week. Maybe it reached a harbor there?
@corneliusroemer @InfrPopGen

[krosa1910]

I think this proposal is linked with the designation of EG.6/EG.6.1. @corneliusroemer could you add the tag to this clade to make it clear

[FedeGueli]

Designated EG.6.1 via d0acddd
cc @InfrPopGen

@krosa1910 could younthe next time put the main mutation in the headline. I scrolled back to search this one but missed to find it

[FedeGueli]

@BorisUitham the third lineage here is still ongoing?

in that case please pre propose it in the other page.

[krosa1910]

Designated EG.6.1 via d0acddd cc @InfrPopGen

@krosa1910 could younthe next time put the main mutation in the headline. I scrolled back to search this one but missed to find it

At the time I mentioned this，EG.5 is a well known concept （still is I would say）and I thought using it as reference for S：456L would be nice. I would usually use main S mutation followed by orf ones now.