BUG: gtdb taxonomy parser (#191)

bokulich-lab · Jun 10, 2024 · 4bf4576 · 4bf4576
1 parent 178cd21
commit 4bf4576
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 2 deletions.
diff --git a/rescript/get_gtdb.py b/rescript/get_gtdb.py
@@ -96,7 +96,7 @@ def _assemble_queries(version='220.0',
 
 
 def parse_gtdb_taxonomy(tax_str):
-    tax = tax_str.split()[0]
+    tax = ' '.join(tax_str.split()[0:2])
     return tax
 
 

diff --git a/rescript/tests/test_get_gtdb.py b/rescript/tests/test_get_gtdb.py
@@ -9,7 +9,7 @@
 import pkg_resources
 from qiime2.plugin.testing import TestPluginBase
 from qiime2.plugins import rescript
-from rescript.get_gtdb import _assemble_queries
+from rescript.get_gtdb import _assemble_queries, parse_gtdb_taxonomy
 from q2_types.feature_data import (TSVTaxonomyFormat, DNAFASTAFormat)
 
 from urllib.request import urlopen
@@ -155,3 +155,12 @@ def _makey_fakey_bact(faking_ignore_this):
                 version='214.1', db_type='All')
             self.assertEqual(str(resc[0].type), 'FeatureData[Taxonomy]')
             self.assertEqual(str(resc[1].type), 'FeatureData[Sequence]')
+
+    def test_parse_gtdb_taxonomy(self):
+        tax_in = ('d__Bacteria;p__Bacillota;c__Bacilli;o__Lactobacillales;'
+                  'f__Lactobacillaceae;g__Oenococcus;s__Oenococcus oeni '
+                  '[locus_tag=NZ_AQVA01000009.1] [location=77871..79431] '
+                  '[ssu_len=1561] [contig_len=79790]')
+        exp = ('d__Bacteria;p__Bacillota;c__Bacilli;o__Lactobacillales;'
+               'f__Lactobacillaceae;g__Oenococcus;s__Oenococcus oeni')
+        self.assertEqual(parse_gtdb_taxonomy(tax_in), exp)