[med-svn] [Git][med-team/q2-phylogeny][upstream] New upstream version 2022.2.0
Andreas Tille (@tille)
gitlab at salsa.debian.org
Wed Jul 20 15:17:07 BST 2022
Andreas Tille pushed to branch upstream at Debian Med / q2-phylogeny
Commits:
1c29f148 by Andreas Tille at 2022-07-20T14:54:54+02:00
New upstream version 2022.2.0
- - - - -
25 changed files:
- .github/workflows/ci.yml
- LICENSE
- − ci/recipe/conda_build_config.yaml
- ci/recipe/meta.yaml
- q2_phylogeny/__init__.py
- q2_phylogeny/_align_to_tree_mafft_fasttree.py
- q2_phylogeny/_align_to_tree_mafft_iqtree.py
- q2_phylogeny/_align_to_tree_mafft_raxml.py
- q2_phylogeny/_fasttree.py
- q2_phylogeny/_filter.py
- q2_phylogeny/_iqtree.py
- q2_phylogeny/_raxml.py
- q2_phylogeny/_util.py
- q2_phylogeny/_version.py
- q2_phylogeny/plugin_setup.py
- q2_phylogeny/tests/__init__.py
- q2_phylogeny/tests/test_align_to_tree_mafft_fasttree.py
- q2_phylogeny/tests/test_align_to_tree_mafft_iqtree.py
- q2_phylogeny/tests/test_align_to_tree_mafft_raxml.py
- q2_phylogeny/tests/test_fasttree.py
- q2_phylogeny/tests/test_filter.py
- q2_phylogeny/tests/test_iqtree.py
- q2_phylogeny/tests/test_raxml.py
- q2_phylogeny/tests/test_util.py
- setup.py
Changes:
=====================================
.github/workflows/ci.yml
=====================================
@@ -50,6 +50,6 @@ jobs:
- uses: qiime2/action-library-packaging at alpha1
with:
package-name: q2-phylogeny
- build-target: tested
+ build-target: dev
additional-tests: py.test --pyargs q2_phylogeny
library-token: ${{ secrets.LIBRARY_TOKEN }}
=====================================
LICENSE
=====================================
@@ -1,6 +1,6 @@
BSD 3-Clause License
-Copyright (c) 2016-2021, QIIME 2 development team.
+Copyright (c) 2016-2022, QIIME 2 development team.
All rights reserved.
Redistribution and use in source and binary forms, with or without
=====================================
ci/recipe/conda_build_config.yaml deleted
=====================================
@@ -1,2 +0,0 @@
-python:
- - 3.8
=====================================
ci/recipe/meta.yaml
=====================================
@@ -1,6 +1,5 @@
{% set data = load_setup_py_data() %}
{% set version = data.get('version') or 'placehold' %}
-{% set release = '.'.join(version.split('.')[:2]) %}
package:
name: q2-phylogeny
@@ -20,14 +19,23 @@ requirements:
run:
- python {{ python }}
- scikit-bio
- - qiime2 {{ release }}.*
- - q2-types {{ release }}.*
- - q2-alignment {{ release }}.*
- - fasttree 2.1.10=0
+ - qiime2 {{ qiime2_epoch }}.*
+ - q2-types {{ qiime2_epoch }}.*
+ - q2-alignment {{ qiime2_epoch }}.*
+ # fasttree pinned because newer versions don't include fasttreemp on osx:
+ # https://github.com/bioconda/bioconda-recipes/blob/master/recipes/fasttree/build.sh
+ - fasttree ==2.1.10=0
- raxml
- iqtree >=1.6.4
+ - pytest
test:
+ requires:
+ - qiime2 >={{ qiime2 }}
+ - q2-types >={{ q2_types }}
+ - q2-alignment >={{ q2_alignment }}
+ - pytest
+
imports:
- q2_phylogeny
- qiime2.plugins.phylogeny
=====================================
q2_phylogeny/__init__.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -10,7 +10,7 @@ from ._util import midpoint_root, robinson_foulds
from ._fasttree import fasttree
from ._raxml import raxml, raxml_rapid_bootstrap
from ._iqtree import iqtree, iqtree_ultrafast_bootstrap
-from ._filter import filter_table
+from ._filter import filter_table, filter_tree
from ._version import get_versions
from ._align_to_tree_mafft_fasttree import align_to_tree_mafft_fasttree
from ._align_to_tree_mafft_iqtree import align_to_tree_mafft_iqtree
@@ -22,4 +22,4 @@ del get_versions
__all__ = ["midpoint_root", "fasttree", "align_to_tree_mafft_fasttree",
"raxml", "raxml_rapid_bootstrap", "iqtree", "filter_table",
"iqtree_ultrafast_bootstrap", "align_to_tree_mafft_iqtree",
- "align_to_tree_mafft_raxml", "robinson_foulds"]
+ "align_to_tree_mafft_raxml", "robinson_foulds", 'filter_tree']
=====================================
q2_phylogeny/_align_to_tree_mafft_fasttree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_align_to_tree_mafft_iqtree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_align_to_tree_mafft_raxml.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_fasttree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -38,8 +38,10 @@ def fasttree(alignment: AlignedDNAFASTAFormat,
cmd = ['FastTree']
else:
env = os.environ.copy()
- n_threads = 0 if n_threads == 'auto' else n_threads
- env.update({'OMP_NUM_THREADS': str(n_threads)})
+ if n_threads == 'auto':
+ env.pop('OMP_NUM_THREADS', 0)
+ else:
+ env.update({'OMP_NUM_THREADS': str(n_threads)})
cmd = ['FastTreeMP']
cmd.extend(['-quote', '-nt', aligned_fp])
=====================================
q2_phylogeny/_filter.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -7,6 +7,8 @@
# ----------------------------------------------------------------------------
import biom
+import numpy as np
+import qiime2
import skbio
@@ -19,3 +21,42 @@ def filter_table(table: biom.Table, tree: skbio.TreeNode) -> biom.Table:
ids_to_keep = tip_ids & feature_ids
table.filter(ids_to_keep, axis='observation', inplace=True)
return table
+
+
+def filter_tree(tree: skbio.TreeNode,
+ table: biom.Table = None,
+ metadata: qiime2.Metadata = None,
+ where: str = None,
+ ) -> skbio.TreeNode:
+ """
+ Prunes a phylogenetic tree to match the input ids
+ """
+ # Checks the input metadata
+ if ((table is None) & (metadata is None)):
+ raise ValueError('A feature table, sequences or metadata must be '
+ 'provided for filtering.')
+ filter_refs = [table, metadata]
+ if np.sum([(ref is not None) for ref in filter_refs]).sum() > 1:
+ raise ValueError('Filtering can only be performed using one reference'
+ ' file. Please choose between filtering with a '
+ 'feature table, sequences, or metadata.')
+ if (where is not None) & (metadata is None):
+ raise ValueError("Metadata must be provided if 'where' is specified")
+
+ # Gets the list of IDs to keep
+ if table is not None:
+ ids_to_keep = table.ids(axis='observation')
+ if metadata is not None:
+ ids_to_keep = metadata.get_ids(where)
+
+ # Gets the list of tips
+ tip_ids = set([t.name for t in tree.tips()])
+
+ # Checks for an intersection between ids
+ if not set(tip_ids).issuperset(set(ids_to_keep)):
+ raise ValueError('The ids for filtering must be a subset of '
+ 'the tips in the tree.')
+ sub_tree = tree.shear(ids_to_keep)
+ sub_tree.prune()
+
+ return sub_tree
=====================================
q2_phylogeny/_iqtree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_raxml.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_util.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/_version.py
=====================================
@@ -23,9 +23,9 @@ def get_keywords():
# setup.py/versioneer.py will grep for the variable names, so they must
# each be defined on a line of their own. _version.py will just call
# get_keywords().
- git_refnames = " (tag: 2021.4.0)"
- git_full = "0396f2627d5b25594d68850f070a73645720386b"
- git_date = "2021-04-20 15:27:09 +0000"
+ git_refnames = " (tag: 2022.2.0)"
+ git_full = "527838b0abdd3754b1fd98709e9cb97e262e6059"
+ git_date = "2022-02-18 19:38:00 +0000"
keywords = {"refnames": git_refnames, "full": git_full, "date": git_date}
return keywords
=====================================
q2_phylogeny/plugin_setup.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -7,10 +7,12 @@
# ----------------------------------------------------------------------------
from qiime2.plugin import (Plugin, Citations, Int, Range, Str, Choices, Bool,
- Float, List)
+ Float, List, TypeMatch, Metadata)
from q2_types.tree import Phylogeny, Unrooted, Rooted
from q2_types.feature_data import FeatureData, AlignedSequence, Sequence
-from q2_types.feature_table import FeatureTable, Frequency
+from q2_types.feature_table import (FeatureTable, Frequency,
+ RelativeFrequency, PresenceAbsence,
+ Composition)
from q2_types.distance_matrix import DistanceMatrix
import q2_phylogeny
@@ -400,12 +402,14 @@ plugin.methods.register_function(
deprecated=True
)
+T1 = TypeMatch([Frequency, RelativeFrequency, PresenceAbsence])
+
plugin.methods.register_function(
function=q2_phylogeny.filter_table,
- inputs={'table': FeatureTable[Frequency],
+ inputs={'table': FeatureTable[T1],
'tree': Phylogeny[Rooted | Unrooted]},
parameters={},
- outputs=[('filtered_table', FeatureTable[Frequency])],
+ outputs=[('filtered_table', FeatureTable[T1])],
input_descriptions={
'table': 'Feature table that features should be filtered from.',
'tree': ('Tree where tip identifiers are the feature identifiers that '
@@ -418,6 +422,41 @@ plugin.methods.register_function(
"are not tip identifiers in tree.")
)
+T2 = TypeMatch([Rooted, Unrooted])
+
+T3 = (Frequency | RelativeFrequency | PresenceAbsence | Composition)
+
+plugin.methods.register_function(
+ function=q2_phylogeny.filter_tree,
+ inputs={'tree': Phylogeny[T2],
+ 'table': FeatureTable[T3],
+ },
+ parameters={'metadata': Metadata,
+ 'where': Str
+ },
+ outputs=[('filtered_tree', Phylogeny[T2])],
+ input_descriptions={
+ 'tree': ('Tree that should be filtered'),
+ 'table': ('Feature table which contains the identifier that should be'
+ ' retained in the tree'),
+ },
+ parameter_descriptions={
+ 'metadata': ("Feature metadata to use with the 'where' statement or "
+ "to select tips to be retained. Metadata objects could "
+ "also include FeatureData[Sequence] data types, if, for"
+ "instance, you want to filter to match represenative "
+ "sequencces."),
+ 'where': ('SQLite WHERE clause specifying sample metadata criteria '
+ 'that must be met to be included in the filtered feature '
+ 'table. If not provided, all samples in `metadata` that'
+ ' are also in the feature table will be retained.'),
+ },
+ output_descriptions={'filtered_tree': 'The resulting phylogenetic tree.'},
+ name="Remove features from tree based on metadata",
+ description=("Remove tips from a tree if their identifiers based on a "
+ "set of provided identifiers.")
+)
+
plugin.methods.register_function(
function=q2_phylogeny.robinson_foulds,
inputs={'trees': List[Phylogeny[Rooted | Unrooted]]},
=====================================
q2_phylogeny/tests/__init__.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_align_to_tree_mafft_fasttree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_align_to_tree_mafft_iqtree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_align_to_tree_mafft_raxml.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_fasttree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -52,7 +52,7 @@ class FastTreeTests(TestPluginBase):
tip_names.sort()
self.assertEqual(tip_names, ['_s_e_q_1_', '_s_e_q_2_'])
- def test_fasttree_n_threads(self):
+ def test_fasttree_n_tips(self):
input_fp = self.get_data_path('aligned-dna-sequences-1.fasta')
input_sequences = AlignedDNAFASTAFormat(input_fp, mode='r')
with redirected_stdio(stderr=os.devnull):
@@ -68,6 +68,51 @@ class FastTreeTests(TestPluginBase):
self.assertEqual(tip_names, ['seq1', 'seq2'])
+# We are using pytest vs. unittest for this test
+# The capfd arg below is a built-in pytest fixture that allows for
+# captured stderr from all subprocesses that directly write
+# to operating system level output
+def test_fasttree_num_threads(capfd):
+ # Rather than providing an actual filepath, we just run a 'help' command
+ # Ideally, we supply --help in this param, but not available in FastTree
+ # So we are using -expert instead, so that FastTree doesn't fail
+ # ######################
+ # Output preview/example:
+ # ######################
+ # Detailed usage for FastTree 2.1.10 Double precision (No SSE3):
+ # FastTree [-nt] [-n 100] [-quote] [-pseudo | -pseudo 1.0]
+ # [-boot 1000 | -nosupport]
+ # [-intree starting_trees_file | -intree1 starting_tree_file]
+ # [-quiet | -nopr]
+ # [-nni 10] [-spr 2] [-noml | -mllen | -mlnni 10]
+ # [-mlacc 2] [-cat 20 | -nocat] [-gamma]
+ # [-slow | -fastest] [-2nd | -no2nd] [-slownni] [-seed 1253]
+ # [-top | -notop] [-topm 1.0 [-close 0.75] [-refresh 0.8]]
+ # [-matrix Matrix | -nomatrix] [-nj | -bionj]
+ # [-lg] [-wag] [-nt] [-gtr] [-gtrrates ac ag at cg ct gt]
+ # [-gtrfreq A C G T]
+ # [ -constraints constraintAlignment
+ # [ -constraintWeight 100.0 ] ]
+ # [-log logfile]
+ # [ alignment_file ]
+ # [ -out output_newick_file | > newick_tree]
+ fasttree('-expert', n_threads=1)
+ captured = capfd.readouterr()
+ assert 'OpenMP' not in captured.err
+
+ fasttree('-expert', n_threads=20)
+ captured = capfd.readouterr()
+ assert 'OpenMP (20 threads)' in captured.err
+
+ # This test case ensures that when a user enters 'auto', the n_threads
+ # var will still be set to the max available on their machine, even if
+ # the OMP_NUM_THREADS env var has been set on their machine
+ os.environ['OMP_NUM_THREADS'] = '2560'
+ fasttree('-expert', n_threads='auto')
+ captured = capfd.readouterr()
+ assert 'OpenMP (2560 threads)' not in captured.err
+
+
class RunCommandTests(TestPluginBase):
package = 'q2_phylogeny.tests'
=====================================
q2_phylogeny/tests/test_filter.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
@@ -10,10 +10,13 @@ import unittest
import io
import numpy as np
+import pandas as pd
from biom.table import Table
import skbio
-from q2_phylogeny import filter_table
+from qiime2 import Metadata
+
+from q2_phylogeny import (filter_table, filter_tree)
class FilterTableTests(unittest.TestCase):
@@ -49,5 +52,76 @@ class FilterTableTests(unittest.TestCase):
self.assertEqual(actual, expected)
+class FilterTreeTests(unittest.TestCase):
+ def setUp(self):
+ rooted_nwk = io.StringIO("((A:0.1, B:0.2)C:0.3, D:0.4, E:0.5)root;")
+ self.tree = skbio.TreeNode.read(rooted_nwk)
+ self.metadata = Metadata(pd.DataFrame(
+ data=np.array([['Bacteria', '1'],
+ ['Archea', '1']], dtype=object),
+ index=pd.Index(['A', 'D'], name='Feature ID'),
+ columns=['kingdom', 'keep'],
+ ))
+ self.table = Table(data=np.array([[0, 1, 2], [2, 2, 2]]),
+ observation_ids=['A', 'D'],
+ sample_ids=['S1', 'S2', 'S3']
+ )
+ self.filtered_tree = self.tree.copy().shear(['A', 'D'])
+ self.filtered_tree.prune()
+
+ def test_filter_tree_error_no_filter_art(self):
+ with self.assertRaises(ValueError) as err:
+ filter_tree(self.tree)
+ self.assertEqual(
+ str(err.exception),
+ ('A feature table, sequences or metadata must be provided for '
+ 'filtering.')
+ )
+
+ def test_filter_tree_error_multiple_filter_arts(self):
+ with self.assertRaises(ValueError) as err:
+ filter_tree(self.tree,
+ table=self.table,
+ metadata=self.metadata)
+ self.assertEqual(
+ str(err.exception),
+ ('Filtering can only be performed using one reference file. '
+ 'Please choose between filtering with a feature table, '
+ 'sequences, or metadata.')
+ )
+
+ def test_filter_tree_error_where_no_metadata(self):
+ with self.assertRaises(ValueError) as err:
+ filter_tree(self.tree,
+ table=self.table,
+ where='[kingdom]="Archea"')
+ self.assertEqual(
+ str(err.exception),
+ ("Metadata must be provided if 'where' is specified")
+ )
+
+ def test_filter_tree_error_filter_superset(self):
+ metadata = Metadata(pd.DataFrame(
+ data=np.array([[1, 1, 0]]).T,
+ index=pd.Index(['A', 'D', 'F'], name='feature-id'),
+ columns=['keep']
+ ))
+ with self.assertRaises(ValueError) as err:
+ filter_tree(self.tree,
+ metadata=metadata)
+ self.assertEqual(
+ str(err.exception),
+ 'The ids for filtering must be a subset of the tips in the tree.'
+ )
+
+ def test_filter_tree_metadata(self):
+ test = filter_tree(self.tree, metadata=self.metadata)
+ self.assertEqual(str(test), str(self.filtered_tree))
+
+ def test_filter_tree_table(self):
+ test = filter_tree(self.tree, table=self.table)
+ self.assertEqual(str(test), str(self.filtered_tree))
+
+
if __name__ == "__main__":
unittest.main()
=====================================
q2_phylogeny/tests/test_iqtree.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_raxml.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
q2_phylogeny/tests/test_util.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
=====================================
setup.py
=====================================
@@ -1,5 +1,5 @@
# ----------------------------------------------------------------------------
-# Copyright (c) 2016-2021, QIIME 2 development team.
+# Copyright (c) 2016-2022, QIIME 2 development team.
#
# Distributed under the terms of the Modified BSD License.
#
View it on GitLab: https://salsa.debian.org/med-team/q2-phylogeny/-/commit/1c29f1488a498f6f0cd2a94b11e7860dc819f1de
--
View it on GitLab: https://salsa.debian.org/med-team/q2-phylogeny/-/commit/1c29f1488a498f6f0cd2a94b11e7860dc819f1de
You're receiving this email because of your account on salsa.debian.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://alioth-lists.debian.net/pipermail/debian-med-commit/attachments/20220720/5361b357/attachment-0001.htm>
More information about the debian-med-commit
mailing list