ENH: Semantic Type for NCBI protein-taxonomy data

Sann5 commented 1 year ago

What's new

Closes #66
Semantic type for a directory containing tree files
- nodes.dmp
- names.dmp
- prot.accession2taxid.gz
Why is this useful? This Semantic Type can be used as optional input to the build-custom-diamond-db action, thereby allowing the user to obtain a database with taxonomy features.

Run it locally

Assuming you have a working environment, run the following.

First, clone the repo and checkout the PR branch:

git clone git@github.com:bokulich-lab/q2-types-genomics.git
gh pr checkout 67

Now let's get you the data to try out the Semantic Type.
```
cd wherever_you_want_to_download_the_data_to
```

Be aware that the next command will download ~15 Gb of data

mkdir ncbi_tax_data
wget ftp://ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdmp.zip -P ncbi_tax_data
wget ftp://ftp.ncbi.nlm.nih.gov/pub/taxonomy/accession2taxid/prot.accession2taxid.gz -P ncbi_tax_data

Now make an artifact with the new semantic type

qiime tools import --input-path ncbi_tax_data --output-path ncbi_tax_data.qza --type ReferenceDB\[TaxonomyNCBI\]

Alternatively, you can run the tests for this Semantic Type (nothing will be downloaded).
```
cd <local_path_to_q2-types-genomics>
pytest
```

codecov[bot] commented 1 year ago

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (22fd9a3) 96.42% compared to head (feb5e4c) 96.56%.

Files	Patch %	Lines
q2_types_genomics/reference_db/_format.py	98.52%	1 Missing :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #67 +/- ## ========================================== + Coverage 96.42% 96.56% +0.13% ========================================== Files 42 42 Lines 1400 1484 +84 ========================================== + Hits 1350 1433 +83 - Misses 50 51 +1 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

Sann5 commented 10 months ago

@misialq

That all LGTM 🏅 I think, though, that this change will require updating the classify-kaiju code in q2-moshpit since you move the format used by that action - could you please double check and maybe open a small PR there to fix that, if needed? Thanks!

I believe there is no update required. The object that is used in classify-kaiju is KaijuDBDirectoryFormat which contains the file formats (NCBITaxonomyNodesFormat and NCBITaxonomyNamesFormat) that were moved with this PR. Now NCBITaxonomyNodesFormat and NCBITaxonomyNamesFormat are imported into the module where KaijuDBDirectoryFormat is defined, instead of being defined in that same module. Only KaijuDBDirectoryFormat is imported by q2-moshpit, not NCBITaxonomyNodesFormat and NCBITaxonomyNamesFormat, so I don't see anything that breaks.

bokulich-lab / q2-types-genomics