Ete4 #750 #751
Ete4 #750 #751
Conversation
There was a problem hiding this comment.
Thanks @dengzq1234
Before I merge this PR, I'd like to understand the couple of things I mentioned before, and also we should understand why after this change the ncbiquery test fails (which seems to be from your change in line 135 of tests/test_ncbiquery.py, that I imagine you had a reason to do):
$ pytest tests/test_ncbiquery.py
[...]
tests/test_ncbiquery.py ......F.
[...]
____________________________________________ Test_ncbiquery.test_merged_id _____________________________________________
self = <tests.test_ncbiquery.Test_ncbiquery testMethod=test_merged_id>
def test_merged_id(self):
ncbi = NCBITaxa(dbfile=DATABASE_PATH)
t1 = ncbi.get_lineage(649756)
> self.assertEqual(t1, [1, 131567, 2, 1783272, 1239, 186801, 3085636, 186803, 207244, 649756])
E AssertionError: Lists differ: [1, 131567, 2, 1783272, 1239, 186801, 186802, 186803, 207244, 649756] != [1, 131567, 2, 1783272, 1239, 186801, 3085636, 186803, 207244, 649756]
E
E First differing element 6:
E 186802
E 3085636
E
E - [1, 131567, 2, 1783272, 1239, 186801, 186802, 186803, 207244, 649756]
E ? ^ ^^^
E
E + [1, 131567, 2, 1783272, 1239, 186801, 3085636, 186803, 207244, 649756]
E ? ^^ + ^^
test_ncbiquery.py:135: AssertionError| def get_rank(self, taxids): | ||
| def _dirty_id_suffix(self, taxid): | ||
| pass | ||
|
|
There was a problem hiding this comment.
I don't see this function used. Why is it defined?
There was a problem hiding this comment.
This is a function I want to do to handle GTDB accesssion,
because I realized sometime they use ids in both way, for example:
GCA_000003645.1 GB_GCA_000003645.1
GCF_900113245.1 RS_GCF_900113245.1
they are equivalent so I want to have a function to handle the fuzzy. But I will keep it in my dev repo
ete4/gtdb_taxonomy/gtdbquery.py
Outdated
| def _get_rank(self, taxids): | ||
| """Return dictionary converting taxids to their GTDB taxonomy rank.""" | ||
| ids = ','.join('"%s"' % v for v in set(taxids) - {None, ''}) | ||
| result = self.db.execute('SELECT taxid, rank FROM species WHERE taxid IN (%s)' % ids) | ||
| return {tax: spname for tax, spname in result.fetchall()} | ||
|
|
||
| def get_rank(self, taxids): | ||
| taxid2rank = {} | ||
| name2ids = self._get_name_translator(taxids) | ||
| overlap_ids = name2ids.values() | ||
| taxids = [item for sublist in overlap_ids for item in sublist] | ||
| """Return dictionary converting taxids to their GTDB taxonomy rank.""" | ||
| ids = ','.join('"%s"' % v for v in set(taxids) - {None, ''}) | ||
| result = self.db.execute('SELECT taxid, rank FROM species WHERE taxid IN (%s)' % ids) | ||
| for tax, rank in result.fetchall(): | ||
| taxid2rank[list(self._get_taxid_translator([tax]).values())[0]] = rank | ||
|
|
||
| return taxid2rank |
There was a problem hiding this comment.
I don't understand this change. Why do we have get_rank() and _get_rank(). What's their difference? Also, get_rank() would need a docstring, and we should avoid having similarly-named functions since that makes it hard to guess which one is appropriate to use.
I changed it because I updated my ncbi taxonomy database, I wonder if it would effect |
|
@jordibc Already update with the next ncbi unitest |
| def get_rank(self, taxids): | ||
| def _dirty_id_suffix(self, taxid): | ||
| pass | ||
|
|
this PR is to solve #750
It includes the following features:
GTDBTaxa()which involves meaningless numeric ids to internal methods, such as:get_lineage_translator()->_get_lineage_translator()get_name_translator()->_get_name_translator()translate_to_names()->_translate_to_names()get_rank()from numeric id to string id inGTDBTaxa(), for example:ignore_unclassifiedin bothNCBITaxa()andGTDBTaxa(), whenignore_unclassified=True,annotate_tree()will igore empty annotation of leaves