XBB.1.5 + S:S494P from multiple states in USA (12 sequences)

[rquiroga7]

Sub-lineage of: XBB.1.5
Earliest sequence: 2022-12-19
Most recent sequence: 2023-01-22
Countries circulating: USA (states: New York, Rhode Island, Connecticut, North Carolina, Massachusetts, Maryland, Florida)

The proposed lineage has ORF8:G8*, and S:494P. I am including the USHER tree although for some reason it shows F486L, although GISAID has F486P for all these sequences. It also makes a mistake with 417 mutations, since all the sequences I am proposing have S:K417N.

Genomes:

USA/MA-CDCBI-CRSP_IKXMYH37FEBGGBBQ/2023|OQ346588.1|2023-01-14
USA/NY-CDC-LC0990938/2023|OQ307674.1|2023-01-14
USA/NY-CDC-LC0978047/2022|OQ205374.1|2022-12-25
USA/RI-CDC-LC0992381/2023|OQ345257.1|2023-01-15
USA/NY-CDC-LC0992366/2023|OQ345214.1|2023-01-15
USA/NY-PRL-221221_00F15/2022|EPI_ISL_16343447|2022-12-19
USA/NY-ASC-210962515/2023|EPI_ISL_16576398|2023-01-03
USA/FL-CDC-LC0977122/2022|OQ206188.1|2022-12-29
USA/MD-HP43388-PIDIGBSHND/2023|EPI_ISL_16674195|2023-01-07
USA/MD-HP43230-PIDRNALJWU/2022|EPI_ISL_16674093|2022-12-29

Also, GISAID finds two additional, newer genomes for a total of 12 sequences:
EPI_ISL_16343447
EPI_ISL_16572620
EPI_ISL_16573542
EPI_ISL_16674093
EPI_ISL_16576398
EPI_ISL_16674195
EPI_ISL_16668777
EPI_ISL_16713029
EPI_ISL_16710218
EPI_ISL_16710261
EPI_ISL_16742093
EPI_ISL_16763385

Usher tree:

[AngieHinrichs]

I am including the USHER tree although for some reason it shows F486L, although GISAID has F486P for all these sequences.

Sorry, there is a bug in the UShER web interface's generation of JSON files to display in nextstrain: mutations (like T23018C) are considered one at a time, even when there is more than one mutation in a codon (like both T23018C and T23019C).

[rquiroga7]

Thank you @AngieHinrichs , that explains a lot!

[corneliusroemer]

ORF8:G8* is always part of XBB.1.5, and S:417N is also normal (if it's K that's most likely an artefact)

It's good to include as many context samples as possible in an Usher tree, yours is a bit sparse making it hard to see if the lineage is clean and where exactly it embeds.

Usher tells me your 12 sequences are part of 3 separate lineages:
https://next.nextstrain.org/fetch/genome-test.gi.ucsc.edu/trash/ct/subtreeAuspice1_genome_test_1b5ef_ca1910.json?f_userOrOld=uploaded%20sample&label=id:node_7948064
https://next.nextstrain.org/fetch/genome-test.gi.ucsc.edu/trash/ct/subtreeAuspice2_genome_test_1b5ef_ca1910.json
https://next.nextstrain.org/fetch/genome-test.gi.ucsc.edu/trash/ct/subtreeAuspice3_genome_test_1b5ef_ca1910.json?f_userOrOld=uploaded%20sample

The biggest one has 9 sequences at the moment - probably best to wait a bit more to see which one grows as we otherwise end up with 100s of lineages.

[rquiroga7]

I think the 3 separate Usher subtrees might be wrong, since the separations are due to what I think are a wrong 486L node and a wrong 417N+417K node.

But I agree waiting a bit is reasonable. Thank you Cornelius

[AngieHinrichs]

Looking at the full tree in taxonium, I think it's likely that the 3 branches should be separate. There are 144 sequences that have C22945T and A1230G like USA/MA-CDCBI-CRSP_IKXMYH37FEBGGBBQ/2023|OQ346588.1 (and then it adds T23042C/S:494P). Then there are more than 7,000 sequences that have T10204C, of which 189 sequences also have A24730T like USA/RI-CDC-LC0993565/2023|OQ353425.1, USA/NY-CDC-LC0978047/2022|OQ205374.1 and USA/NY-CDC-LC0990938/2023|OQ307674.1.

There's even a fourth XBB.1.5 > T17124C > ... > T23042C/S:494P branch in the 2023-02-02 tree: 26 sequences have C8655T, then one sequence has C8655T and G23628T (S:S689I), and then USA/PA-CDC-QDX45433024/2023|OQ329235.1|2023-01-09 adds T23042C/S:494P.

Like @corneliusroemer said it's helpful to use a larger subtree size. The UShER web interface's default number of samples per subtree (50) was set with little local outbreaks in mind. For analyzing potential lineages, it's better to use a much larger size like 1000 or 5000 in order to get as much context as possible.

[rquiroga7]

Thank you for the input! Seems like S494P is actually appearing on multiple different lineages. Thank you also for the feedback regarding subtree sample number. Will keep that in mind.

[rquiroga7]

There are now 37 XBB1.5* sequences with S494P in GISAID, but as @AngieHinrichs suggested, they appear to be from many separate branches. Now also appearing in England and Denmark.

[corneliusroemer]

I'll close this due to homoplasy - if there's a single big S:S494P, please open an issue for it specifically