Merged
Conversation
This was referenced Mar 22, 2026
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
4ca0213 to
4786af7
Compare
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Mar 31, 2026
Augur PR <nextstrain/augur#1958> Auspice PR TODO XXX
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
4786af7 to
242ced3
Compare
jameshadfield
commented
Apr 1, 2026
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Apr 1, 2026
Augur PR <nextstrain/augur#1958> Auspice PR TODO XXX
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
242ced3 to
889a7a3
Compare
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #1958 +/- ##
==========================================
+ Coverage 75.06% 75.11% +0.05%
==========================================
Files 83 83
Lines 9708 9810 +102
Branches 1969 1998 +29
==========================================
+ Hits 7287 7369 +82
- Misses 2084 2096 +12
- Partials 337 345 +8 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
jameshadfield
added a commit
to nextstrain/auspice
that referenced
this pull request
Apr 1, 2026
Follows the work done in Augur <nextstrain/augur#1958> which relaxes the JSON schema to allow annotations without a `nuc` entry. We create a placeholder one in Auspice which spans the observed CDS (nuc) ranges so that the entropy rendering code can still operate.
3 tasks
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Apr 2, 2026
Augur PR <nextstrain/augur#1958> Auspice PR <nextstrain/auspice#2040>
joverlee521
requested changes
Apr 3, 2026
Contributor
joverlee521
left a comment
joverlee521
left a comment
There was a problem hiding this comment.
Left general design questions, but the only blocking comment is to verify single gene vs multiple genes defined in a single file.
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Apr 7, 2026
Augur PR <nextstrain/augur#1958> Auspice PR <nextstrain/auspice#2040>
jameshadfield
force-pushed
the
james/ancestral-bug-fixes
branch
from
a0c8197 to
d1d7522
Compare
jameshadfield
added a commit
that referenced
this pull request
Apr 14, 2026
Review feedback by @joverlee521 <#1958 (comment)> Using an explicit argument when the root sequence represents proteins improves both code clarity and (self) documentation.
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
889a7a3 to
72dad75
Compare
Allows for protein-only analyses to be exportable. Note that Auspice PR <nextstrain/auspice#2040> is needed for the entropy panel to display for such datasets.
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Apr 14, 2026
Augur PR <nextstrain/augur#1958> Auspice PR <nextstrain/auspice#2040>
jameshadfield
added a commit
that referenced
this pull request
Apr 14, 2026
Review feedback by @joverlee521 <#1958 (comment)> Using an explicit argument when the root sequence represents proteins improves both code clarity and (self) documentation.
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
72dad75 to
db37bb1
Compare
Member
Author
joverlee521
requested changes
Apr 14, 2026
Contributor
joverlee521
left a comment
joverlee521
left a comment
There was a problem hiding this comment.
Thanks for going through the other augur commands (sorry I opened the rabbit hole)! My main concern is the auto-detect of sequence types in augur index can potentially change behavior with the sequence filters.
| sequence_filter_group.add_argument('--min-length', type=int, help=descriptions['min_length']) | ||
| sequence_filter_group.add_argument('--max-length', type=int, help=descriptions['max_length']) | ||
| sequence_filter_group.add_argument('--non-nucleotide', action='store_true', help=descriptions['non_nucleotide']) | ||
| sequence_filter_group.add_argument('--exclude-invalid', action='store_true', help=descriptions['exclude_invalid']) |
Contributor
There was a problem hiding this comment.
Related to my other comment for index, thoughts on adding a --seq-type option here to be able to override auto-detect.
If I somehow have a bad nucleotide sequence "AAAETCGG", I'd expect it to be filtered out by --exclude-invalid, but in this case it would auto-detected as "aa" and pass the filter.
Contributor
There was a problem hiding this comment.
Pushed up 00d5223 to show example Cram test that passes on master, but not on this branch.
Member
Author
There was a problem hiding this comment.
If we go with --seq-type I'll change refine's proposed --aa flag to --seq-type aa, yeah?
Contributor
There was a problem hiding this comment.
Oh huh, yeah probably best to match the options across commands. So I guess there's two choices here
- Keep auto-detect for index/filter. All commands support
--seq-type aa/nucto override. - No auto-detect at all. All commands default to
nucand support the--aaflag to switch toaa.
We could also add a new envvar AUGUR_SEQ_TYPE to set the seq type across multiple commands.
Member
Author
There was a problem hiding this comment.
As discussed in dev chat, moved to --seq-type throughout. We can add AUGUR_SEQ_TYPE later as desired.
victorlin
reviewed
Apr 15, 2026
TreeTime already supported (one!) AA model, so all that was required on the Augur side was adding an "--seq-type" arg and changing the TreeTime parameterisation accordingly.
Adds a fair bit of code complexity but one step closer to augur pipelines which are protein-only. The actual reconstruction is unchanged, as we were already doing independent nuc / aa reconstructions. There are some rough edges to be tidied up in subsequent commits: 1. The annotation parsing requires the file to define the nucleotide coordinates, which we don't need for a protein-only analysis. Allowing the nucleodide definition to be optional is somewhat complicated, and most annotations files will have them anyways. However if we're only reconstructing a single gene it'd be really nice to make the annotations file optional, we can make up dummy nuc coordinates for the gene (e.g. 1..3*aa_len) easily. 2. Allow a (AA) root-sequence to be provided. There are different ways to do this, but the simplest would be to allow them to be provided analogously to the ``--translations`` argument.
(if we are only reconstructing amino-acid sequences). For a typical single-gene amino-acid analysis it would often be frustrating to need to create a genemap when we can create dummy nucleotide coordinates to represent the gene and the interpretation of result will be just as valid.
Review feedback by @joverlee521 <#1958 (comment)> Using an explicit argument when the root sequence represents proteins improves both code clarity and (self) documentation.
jameshadfield
added a commit
that referenced
this pull request
Apr 20, 2026
This was motivated by the desire to allow sequence-based filters for `augur filter`, which will require the index to be run for AA sequences. The choice to use an explicit `--seq-type` argument came from PR discussion <#1958 (comment)>
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
db37bb1 to
e6a346c
Compare
jameshadfield
added a commit
to nextstrain/docs.nextstrain.org
that referenced
this pull request
Apr 20, 2026
Augur PR <nextstrain/augur#1958> Auspice PR <nextstrain/auspice#2040>
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
e6a346c to
e93b48e
Compare
This was motivated by the desire to allow sequence-based filters for `augur filter`, which will require the index to be run for AA sequences. The choice to use an explicit `--seq-type` argument came from PR discussion <#1958 (comment)>
explicitly to check behaviour of whether non-A/T/G/C characters count towards the length (they don't - good!)
The only change for nucleotide sequences (i.e. every existing usage) is the argument change of '--non-nucleotide' to '--exclude-invalid'. (I think the new argument is more self-explanatory!)
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
e93b48e to
4280dc2
Compare
Deprecates the 'non_nucleotide' key (see parent commit)
jameshadfield
force-pushed
the
amino-acid-augur
branch
from
4280dc2 to
e6a6972
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This allows protein sequences to be analysed in a typical augur pipeline (align, tree, refine, ancestral, traits, export)
I've created this development workflow in seasonal-flu to run real-life AA-only analyses for testing purposes. I have a PR on docs.nextstrain.org with a guide on how to run such analyses.
See commit messages for details
This PR is on top of #1975
This will close #820 but needs to be paired with new docs and a new auspice version
To Do
XnotNfor aa seqs (#522) neherlab/treetime#523 is merged and a new treetime release made, then enforce a minimum treetime version for protein analyses (not actually needed, our ancestral reconstruction is being done withJC69(alphabet=aa))