New JN.1.18 FLiRT lineage with S:L492L and Orf6:I60T likely emerged in Africa (32) with one S:K182I sublineage (9 seqs), one with S:A222V (5, NZL)

[FedeGueli]

Original proposal
Formerly tracked as Branch 87 of sars-cov-2-variants/lineage-proposals#1089

JN.1.18 (S:R346T)> S:F456L (T22930G), T23036C (Silent:S:L492L) , Orf6:I60T (T27380C) thx @aviczhl2 for noting the right placement .
Query: T27380C,T22930G
Samples: 5 ( 6 on Usher) with one sample from Ethiopia and one from Ghana
EPI_ISL_19029565, EPI_ISL_19062760, EPI_ISL_19070234,
EPI_ISL_19075644, EPI_ISL_19075660
I want to highlight that silent mutation S:L492L (T23036C) could open up a way to 492P/Q/R if 23037 will mutate further

Tree:

There are two branches:
One with S:K182I
[defining #2567 >>Orf1a:G604S (G2075A), S:K182I (A22107T) ]
query: G2075A, A22107T, T27380C
Samples: 3

One with S:I197V (A22151G), S:P251H
#2567 > ORF1a:E1766A (A5562C),C9565T, T16950C S:I197V (A22151G), S:P251H (C22314A), C27247T
Query: C9565T, T16950C ,A22151G
Samples: 2

One with S:F1103S
#2567 > S:F1103S (T24870C)
Query: T24870C,T27380C
Samples: 3 (NY), (Ca)

One with S:A222V
#2567 > S:A222V (C22227T)
Query : C22227T, T27380C,T22930G
Samples: 5 (NZL)

[FedeGueli]

8 now , the last two GBW (likely same patient) samples have two additional spike mutations: S:I197V and S:P251H

[FedeGueli]

10 growing

[FedeGueli]

@AngieHinrichs now this has been misplaced under the Branch 20 of sars-cov-2-variants/lineage-proposals#1253 but i think it is very hard to think it belongs really there cause this has T22930G while Branch 20 is the collector of everything with T22930A ( in the same way Jn.1.16 did for T22928C)

cc @corneliusroemer @aviczhl2

[aviczhl2]

The 27143 branch is errorous and usher has fixed it. This one shall be sub-branch of JN.1.18

[FedeGueli]

The 27143 branch is errorous and usher has fixed it. This one shall be sub-branch of JN.1.18

Thx!

[AngieHinrichs]

@AngieHinrichs now this has been misplaced under the Branch 20 of sars-cov-2-variants/lineage-proposals#1253 but i think it is very hard to think it belongs really there cause this has T22930G while Branch 20 is the collector of everything with T22930A ( in the same way Jn.1.16 did for T22928C)

Yes, UShER and/or matOptimize is placing it wrong there. Thanks for the reminder, I need to make a simple test case for a bug report.

[FedeGueli]

@AngieHinrichs now this has been misplaced under the Branch 20 of sars-cov-2-variants/lineage-proposals#1253 but i think it is very hard to think it belongs really there cause this has T22930G while Branch 20 is the collector of everything with T22930A ( in the same way Jn.1.16 did for T22928C)

Yes, UShER and/or matOptimize is placing it wrong there. Thanks for the reminder, I need to make a simple test case for a bug report.

Thank you!

[AngieHinrichs]

When I look at this more closely, changing the tree would not reduce the parsimony penalty. The current structure has 3 mutations on one branch and 1 on another:
T22930A > G22599C > A22930G
G22599C
To our eyes, this structure looks more likely:
T22930A > G22599C
G22599C > T22930G
-- but it is still 4 mutations in total, so from a parsimony point of view it is equivalent. And using parsimony is what makes usher and matOptimize fast enough to keep up with the large size of the tree. So strictly speaking it is hard to cast this as a bug even though it is annoying, and it would be nicer if usher and matOptimize could also look at the effect on path lengths or slightly penalize paths that have successive mutations at the same position.

The good news is that if I force-move the node to where we want it to go, matOptimize should leave it there because it can't improve parsimony by moving it where we don't want it to go. So I will give that a try.

[aviczhl2]

When I look at this more closely, changing the tree would not reduce the parsimony penalty. The current structure has 3 mutations on one branch and 1 on another: T22930A > G22599C > A22930G G22599C To our eyes, this structure looks more likely: T22930A > G22599C G22599C > T22930G -- but it is still 4 mutations in total, so from a parsimony point of view it is equivalent. And using parsimony is what makes usher and matOptimize fast enough to keep up with the large size of the tree. So strictly speaking it is hard to cast this as a bug even though it is annoying, and it would be nicer if usher and matOptimize could also look at the effect on path lengths or slightly penalize paths that have successive mutations at the same position.

The good news is that if I force-move the node to where we want it to go, matOptimize should leave it there because it can't improve parsimony by moving it where we don't want it to go. So I will give that a try.

I think some penalty shall be added for mutating the same position multiple times in short gaps. We(and the system) shall assume things like 22930T->22930A->22930G in a short time is unlikely to occur.

[FedeGueli]

• 5 NZL 1% of prevalence there after April 20 all with S:A222V

[FedeGueli]

24

[AngieHinrichs]

I did a force move of this branch (NewZealand/24ZA2051/2024) to JN.1.18 a couple days ago and so far matOptimize has left it in place.

[FedeGueli]

I did a force move of this branch (NewZealand/24ZA2051/2024) to JN.1.18 a couple days ago and so far matOptimize has left it in place.

Thank you Angie!

[FedeGueli]

Designated JN.1.18.4 ,via cdf43cc
LH.1 (K182I) via 95b07a7
LH.2 (197V , 251H) via 9bc4a49