Skip to content

Commit

Permalink
Designate HK.10 (EG.5.1.1.10, S:L455F, on C19983T branch, China/Engla…
Browse files Browse the repository at this point in the history 
…nd/Denmark) with 19 seqs
  • Loading branch information
@corneliusroemer
corneliusroemer committed Sep 1, 2023
1 parent 36e3fb0 commit f655671
Show file tree
Hide file tree
Showing 2 changed files with 20 additions and 0 deletions.
1 change: 1 addition & 0 deletions lineage_notes.txt
View file
19 changes: 19 additions & 0 deletions lineages.csv
View file

7 comments on commit f655671

@FedeGueli
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@corneliusroemer @AngieHinrichs this is Branch 9 from sars-cov-2-variants/lineage-proposals#537

but @NkRMnZr found an issue withthis lineage the entire danish branch has a weird reversion: T19983C

so designating together most of the sequence of this lineage doesnt have the defining C19983T

see here:sars-cov-2-variants/lineage-proposals#537 (comment)

and here my counter analasys (i didnt find evidence of recombination although a Fe.1.2 circualting in Denmark could have been a candidate):
sars-cov-2-variants/lineage-proposals#537 (comment)

@Over-There-Is
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Not only the Danish sequences, also the British sequence England/CLIMB-CM7YGJST does not have C19983T, it has Ns on 19790~20042. It is reasonable to believe that these sequences are from a HK.3 subbranch. (or descendant of EG.5.1.1> C934T?)
And the Shanghai local sequences seem belonging EG.5.1.1> C6695T> C9448A> C5575T, G22927T> G23006C branch, and I don't know how Shanghai/SCDC-HG0815-3112/2023(imported from Japan) gets C19983T.

@AngieHinrichs
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for pointing out England/CLIMB-CM7YGJST @Over-There-Is! The remove/re-optimize/re-place trick seems to have fixed it -- that little branch is placed in HK.3 as of the 2023-09-07 tree.

@corneliusroemer
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, this is more reason not to designate too much 455 in HK yet as there are just so many artefacts :(

@FedeGueli
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks, this is more reason not to designate too much 455 in HK yet as there are just so many artefacts :(

In reality @AngieHinrichs , uder our expressed request on #537 checked the whole tree and there arent so many , the most evident ones have been identified and requested to a second check. But waiting is not bad just hope it is clear everyone what amount of time is needed to check the whole thing , designating the fastest and the ones reaching to say 50 will help people in the background too doing a superlative work in tracking this combo.

@AngieHinrichs
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To be fair, I only checked the tree and not the sequences themselves (that would take a lot more time). But I believe there really is a lot of true homoplasy (independent mutations followed by enough spread to be picked up by current sequencing) going on.

But here's the other side of my opinion: every time a lineage is designated, it makes a little work for me in that I have to look in taxonium at where the designated sequences land in the tree, identify the specific node that represents the start of the lineage, get the path of mutations to that node relative to its parent lineage, and update pango.clade-mutations.tsv so that the new lineage will be annotated on the tree in the next build. I also have to do some maintenance of pango.clade-mutations.tsv: when a lineage is designated early and then more sequences are added, the order of mutations leading to the lineage may change, and sometimes there are mutations that we care about more than others (like S: changes vs. some random position with no amino acid change), and the specific list of mutations for the start of the lineage needs to be modified. So... sometimes I wish lineages were designated a little less frequently.

But all that does not change my appreciation for the massive amount of work that you all do to discover and track branches with these interesting mutations! Is there a way that we can make you feel appreciated even when a lineage is not designated? There is a public record of your work in these github issues, it's not invisible.

@FedeGueli
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

To be fair, I only checked the tree and not the sequences themselves (that would take a lot more time). But I believe there really is a lot of true homoplasy (independent mutations followed by enough spread to be picked up by current sequencing) going on.

But here's the other side of my opinion: every time a lineage is designated, it makes a little work for me in that I have to look in taxonium at where the designated sequences land in the tree, identify the specific node that represents the start of the lineage, get the path of mutations to that node relative to its parent lineage, and update pango.clade-mutations.tsv so that the new lineage will be annotated on the tree in the next build. I also have to do some maintenance of pango.clade-mutations.tsv: when a lineage is designated early and then more sequences are added, the order of mutations leading to the lineage may change, and sometimes there are mutations that we care about more than others (like S: changes vs. some random position with no amino acid change), and the specific list of mutations for the start of the lineage needs to be modified. So... sometimes I wish lineages were designated a little less frequently.

But all that does not change my appreciation for the massive amount of work that you all do to discover and track branches with these interesting mutations! Is there a way that we can make you feel appreciated even when a lineage is not designated? There is a public record of your work in these github issues, it's not invisible.

Thank you Angie, i m conscious of the great work starting with designation xhoices and scanning of seqs by Cornelius and your one . Our goal to be clear it is not appreciation but to build a solidly described tree and if any to be useful for everyone in the long chain!
The fact here that you have seen with your eyes we have a high level of homoplasy concentrated in a short time ,1month and half? , And very soon after the "mother" lineage started spreading so very little diversity sometimes one nuc only between all these sublineages.
If we agree on a threeshold to be proposed, revieweed and then designated that would be super cool . I arbitrarly fixed one very low to try to anticipate the wave but ok now i have raised it to 50. But being all of them at least as fast as dominant strains it just delays and doesnt solve the thing. Thank you for your patience and work i prefectly understand both your points ,and it is not to bother further just to advice in advance

Please sign in to comment.