Use gene reference files to generate E gene trees #48
Conversation
Until we have a set of E gene serotype and genotype defining mutations, skip the augur clades call for E gene builds. This was originally discussed for commit: 2f48ddb Used in measles: nextstrain/measles@ab92a1b And diagrammed in: #17 (comment)
Add ncbi_serotype annotation to phylogeny, since this can be used as a source-of-truth when adding E gene defining serotypes in subsequent PRs.
Mirroring the change from measles commit: * nextstrain/measles@862dbaf
There was a problem hiding this comment.
Thanks for breaking down the steps @j23414! I've left some non-blocking comments for consideration, but this looks good to merge to me 👍
| @@ -42,7 +42,7 @@ rule prepare_auspice_config: | |||
| output: | |||
| auspice_config="results/config/{gene}/auspice_config_{serotype}.json", | |||
| params: | |||
| replace_clade_key="clade_membership", | |||
| replace_clade_key=lambda wildcard: r"clade_membership" if wildcard.gene in ['genome'] else r"nextclade_subtype", | |||
There was a problem hiding this comment.
Because the clade_membership is automatically used to create the clade branch label in augur export, this change means the E gene build will not have the automatic clade branch label.
It's possible to create custom branch labels since nextstrain/augur#728, but this uses augur clades to create the labels and the workflow is skipping augur clades for the E gene build 😅
There was a problem hiding this comment.
Since the nextclade_subtype field is being added through augur traits, maybe augur traits should be updated to support adding branch labels as well...
There was a problem hiding this comment.
Good to know! I'm currently planning to explore serotype and genotype-defining E mutations (which would create clade_membership for E gene builds) in a future PR.
But if that doesn't work out, I'll explore "create custom branch labels ..." route you've referenced.
| @@ -42,7 +42,7 @@ rule prepare_auspice_config: | |||
| output: | |||
| auspice_config="results/config/{gene}/auspice_config_{serotype}.json", | |||
| params: | |||
| replace_clade_key="clade_membership", | |||
| replace_clade_key=lambda wildcard: r"clade_membership" if wildcard.gene in ['genome'] else r"nextclade_subtype", | |||
| replace_clade_title=lambda wildcard: r"Serotype" if wildcard.serotype in ['all'] else r"DENV genotype", | |||
There was a problem hiding this comment.
To mirror the title NCBI serotype added below, maybe the default clade_membership title should be updated to Nextstrain serotype?
There was a problem hiding this comment.
Thank you for flagging! I'm considering renaming NCBI serotype toSerotype (NCBI) to better match naming convention used in measle's Genotype (NCBI) title. Then, I agree that the default title for clade_membership should be renamed from Serotype to Serotype (Nextstrain).
To recap:
NCBI serotype->Serotype (NCBI): indicating that denv1-4 assignment is based on NCBI GenBank record annotationSerotype->Serotype (Nextstrain): indicating that denv1-4 assignment is based on augur clades call using full-genome-level-serotype-defining amino acid mutationsNextclade genotype->Genotype (Nextclade): indicating genotype level assignment within the serotype (e.g.DENV1/S) based on Nextclade call
This naming adjustment leaves space for a potential Genotype (NCBI) if we develop a script for parsing genotype annotations from the GenBank data.
There was a problem hiding this comment.
Ah, actually I think I'll merge this pull request as it is, and handle the renaming mentioned above in a later PR. That way, we can address it along with fixing this issue: #41.
Description of proposed changes
Following up on the Generating gene reference files PR, use the E gene reference files to create E gene trees.
Visual summary (view whole pipeline plan so far)
Related issue(s)
Checklist
Links to resulting trees below
A manual run of the dengue serotype 4 E gene tree might look like:
# from dengue repo top level nextstrain build phylogenetic auspice/dengue_denv4_E.json