Ingest: fetch and append pathoplexus global lineage calls

[j23414]

Description of proposed changes

This pull request adds rules to retrieve global lineage (1A, 1B, 2, 3, 4, 5, 7, 8) calls from the Pathoplexus website. After appending the Pathoplexus lineages, we can color the tree as shown at https://next.nextstrain.org/staging/WNV/pathoplexus?c=lineage

This change has the added benefit of restricting the USA-based clade calls to the 1A strains (which include most North American WNV strains) since they are based on an initial WNV case isolated in New York from 1999.

Left (original global tree); Right (fixed global tree, where USA-based clade calls restricted to 1A strains)

The updated tree can be viewed at https://next.nextstrain.org/staging/WNV/pathoplexus?c=clade_membership

Related issue(s)

• Related to Perhaps append Pathoplexus WNV lineage calls #35
• Prerequisite for Use lineages for build targets #34
• And solves Restrict Nextclade assignments to NY99 and sub-branches, not across the global tree #25

Checklist

• Checks pass

[tsibley]

FWIW, Pathoplexus is using the Nextclade dataset nextstrain/wnv/all-lineages to make those lineage calls. Not sure if you were aware or not, so thought I'd mention it. From a reliability and robustness perspective, it'd be better to run nextclade ourselves to make those calls rather than pull from Pathoplexus.

[j23414]

Thanks @tsibley! I should have documented that @joverlee521 tracked down the nextstrain/wnv/all-lineages dataset to a draft pull request. Extracting a nextclade dataset from a draft branch seemed fragile, therefore I opted for test driving the public facing Pathoplexus API.

Once the nextstrain/wnv/all-lineages dataset is merged and properly documented, I'm happy to transition to using a nextclade call!

[tsibley]

Ah, ok. Fair enough. I'll note that the dataset is still obtainable with nextclade dataset get and can be pinned to the specific current commit of that PR so that any future changes to the wnv branch aren't automatically pulled in.

nextclade dataset get \
  --name nextstrain/wnv/all-lineages \
  --server https://raw.githubusercontent.com/nextstrain/nextclade_data/8f2e791d3a59013ee88e1d1d7e83b486d39c4ecb/data_output \
  --output-zip wnv.zip

But point taken about it being only in a draft PR; I hadn't clocked that!

[DOH-LMT2303]

Overall the Pathoplexus lineages are in agreement with the topology of the tree and the literature. However, there is need to double check the classification of PP445212 as Pathoplexus lineage 3, because only lineage 1a has been detected in North America at least until the publication of this paper Koch. et al. 2023.

[j23414]

PP445212 is strange! I tracked it down to a mixed lineage 1 and 3 infection in Nebraska reported by the CDC in 2024. As far as I can see, there doesn't seem to be a persistent detection of lineage 3 in North America following that case. 🤞