def test_reconstruct_counts_unweighted(self):
count_table = reconstruct_counts(
region=['Bludhaven', 'Gotham'],
regional_alignment=[
ts.region1_align.view(Delayed).copy(),
ts.region2_align.view(Delayed).copy()
],
regional_table=[
ts.region1_counts.view(biom.Table),
ts.region2_counts.view(biom.Table)
],
database_map=self.datbase_map['clean_name'],
database_summary=self.database_summary,
debug=True,
min_counts=10,
min_abund=1e-2,
region_normalize='unweighted')
npt.assert_array_equal(
count_table.matrix_data.todense(),
np.array([[200, 100, 0, 50, 100, 100], [200, 50, 200, 25, 50, 50],
[0, 200, 200, 0, 100, 100]]).T * 1.)
npt.assert_array_equal(np.array(list(count_table.ids(axis='sample'))),
np.array(['sample1', 'sample2', 'sample3']))
npt.assert_array_equal(
np.array(list(count_table.ids(axis='observation'))),
np.array(['seq1', 'seq2', 'seq3', 'seq4', 'seq5', 'seq6']),
)
def test_reconstruct_counts_align_drop_samples_warning(self):
with warnings.catch_warnings(record=True) as w:
count_table, summary, mapping = reconstruct_counts(
region=['Bludhaven', 'Gotham'],
regional_alignment=[ts.region1_align.view(pd.DataFrame).copy(),
ts.region2_align.view(pd.DataFrame).copy()],
kmer_map=[ts.region1_db_map.view(pd.DataFrame).copy(),
ts.region2_db_map.view(pd.DataFrame).copy()],
regional_table=[ts.region1_counts.view(biom.Table).copy(),
ts.region2_counts.view(biom.Table).copy()],
debug=True,
min_counts=590,
min_abund=1e-2
)
self.assertTrue(issubclass(w[-1].category, UserWarning))
self.assertEqual(str(w[-1].message),
'There are 2 samples with fewer than 590 '
'total reads. These samples will be discarded.')
def test_reconstruct_counts_align_drop_samples_error(self):
with self.assertRaises(ValueError) as err:
count_table, summary, mapping = reconstruct_counts(
region=['Bludhaven', 'Gotham'],
regional_alignment=[ts.region1_align.view(pd.DataFrame).copy(),
ts.region2_align.view(pd.DataFrame).copy()],
kmer_map=[ts.region1_db_map.view(pd.DataFrame).copy(),
ts.region2_db_map.view(pd.DataFrame).copy()],
regional_table=[ts.region1_counts.view(biom.Table).copy(),
ts.region2_counts.view(biom.Table).copy()],
debug=True,
min_abund=1e-2
)
self.assertEqual(
str(err.exception),
'None of the samples have more than the 1000 total '
'sequences required for reconstruction.'
)
def test_reconstruct_counts_align_drop_samples_error(self):
with self.assertRaises(ValueError) as err:
count_table = reconstruct_counts(
region=['Bludhaven', 'Gotham'],
regional_alignment=[
ts.region1_align.view(pd.DataFrame).copy(),
ts.region2_align.view(pd.DataFrame).copy()
],
regional_table=[
ts.region1_counts.view(biom.Table),
ts.region2_counts.view(biom.Table)
],
database_map=self.datbase_map['clean_name'],
database_summary=self.database_summary,
debug=True,
min_counts=590,
min_abund=1e-2)
self.assertEqual(
str(err.exception),
'There are 2 samples with fewer than 590 total sequences. '
'Please check your minimum counts and make sure your '
'representative sequences are aligned with the database.')
def test_reconstruct_counts_same_twice(self):
known_map = pd.DataFrame(
data=[['seq1|seq2', 'WANTCAT', 'WANTCAT', 15],
['seq1|seq2', 'WANTCAT', 'WANTCAT', 15],
['seq3', 'WANTCAT', 'WANTCAT', 15],
['seq5', 'WANTCAT', 'WANTCAT', 15],
['seq6', 'WANTCAT', 'WANTCAT', 15]],
index=pd.Index(['seq1', 'seq2', 'seq3', 'seq5', 'seq6'],
name='db-seq'),
columns=['clean_name', 'first-fwd-primer', 'last-fwd-primer',
'last-kmer-length'],
)
known_summary = pd.DataFrame.from_dict(orient='index', data={
'seq1|seq2': {'num-regions': 1.,
'total-kmers-mapped': 2.,
'mean-kmer-per-region': 2.,
'stdv-kmer-per-region': 0.,
'mapped-asvs': 'asv01'
},
'seq3': {'num-regions': 1,
'total-kmers-mapped': 2,
'mean-kmer-per-region': 2,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv02|asv03'
},
'seq5': {'num-regions': 1,
'total-kmers-mapped': 1,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv04|asv05',
},
'seq6': {'num-regions': 1,
'total-kmers-mapped': 1,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv04|asv05',
},
})
known_summary.index.set_names('feature-id', inplace=True)
count_table, summary, mapping =reconstruct_counts(
region=['Bludhaven', 'Bludhaven'],
regional_alignment=[ts.region1_align.view(pd.DataFrame).copy(),
ts.region1_align.view(pd.DataFrame).copy()],
kmer_map=[ts.region1_db_map.view(pd.DataFrame).copy(),
ts.region1_db_map.view(pd.DataFrame).copy()],
regional_table=[ts.region1_counts.view(biom.Table).copy()],
debug=True,
min_counts=10,
min_abund=1e-2,
region_normalize='unweighted'
)
npt.assert_array_equal(
np.array(list(count_table.ids(axis='sample'))),
np.array(['sample1', 'sample2', 'sample3'])
)
npt.assert_array_equal(
np.array(list(count_table.ids(axis='observation'))),
np.array(['seq1|seq2', 'seq3', 'seq5', 'seq6']),
)
npt.assert_array_equal(
count_table.matrix_data.todense(),
np.array([[150, 0, 50, 50],
[125, 100, 25, 25],
[100, 100, 50, 50]]).T * 1.,
)
pdt.assert_frame_equal(mapping, known_map)
pdt.assert_frame_equal(summary.to_dataframe(),
known_summary)
def test_reconstruct_counts(self):
known_map = pd.DataFrame(
data=[['seq1', 'WANTCAT', 'CACCTCGTN', 15],
['seq2', 'WANTCAT', 'CACCTCGTN', 15],
['seq3', 'WANTCAT', 'CACCTCGTN', 15],
['seq4', 'CACCTCGTN', 'CACCTCGTN', 15],
['seq5', 'WANTCAT', 'CACCTCGTN', 15],
['seq6', 'WANTCAT', 'CACCTCGTN', 15]],
index=pd.Index(['seq1', 'seq2', 'seq3', 'seq4', 'seq5', 'seq6'],
name='db-seq'),
columns=['clean_name', 'first-fwd-primer', 'last-fwd-primer',
'last-kmer-length'],
)
known_summary = pd.DataFrame.from_dict(orient='index', data={
'seq1': {'num-regions': 2.,
'total-kmers-mapped': 2.,
'mean-kmer-per-region': 1.,
'stdv-kmer-per-region': 0.,
'mapped-asvs': 'asv01|asv06'
},
'seq2': {'num-regions': 2,
'total-kmers-mapped': 2,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv01|asv07',
},
'seq3': {'num-regions': 2,
'total-kmers-mapped': 3,
'mean-kmer-per-region': 1.5,
'stdv-kmer-per-region': np.std([1, 2], ddof=1),
'mapped-asvs': 'asv02|asv03|asv08'
},
'seq4': {'num-regions': 1,
'total-kmers-mapped': 1,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv09'
},
'seq5': {'num-regions': 2,
'total-kmers-mapped': 2,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv04|asv05|asv10',
},
'seq6': {'num-regions': 2,
'total-kmers-mapped': 2,
'mean-kmer-per-region': 1,
'stdv-kmer-per-region': 0,
'mapped-asvs': 'asv04|asv05|asv11',
},
})
known_summary.index.set_names('feature-id', inplace=True)
count_table, summary, mapping = reconstruct_counts(
region=['Bludhaven', 'Gotham'],
regional_alignment=[ts.region1_align.view(pd.DataFrame).copy(),
ts.region2_align.view(pd.DataFrame).copy()],
kmer_map=[ts.region1_db_map.view(pd.DataFrame).copy(),
ts.region2_db_map.view(pd.DataFrame).copy()],
regional_table=[ts.region1_counts.view(biom.Table),
ts.region2_counts.view(biom.Table)],
debug=True,
min_counts=10,
min_abund=1e-2)
npt.assert_array_equal(
np.array(count_table.matrix_data.todense()),
np.array([[100, 50, 0, 50, 50, 50],
[100, 25, 100, 25, 25, 25],
[ 0, 100, 100, 0, 50, 50]]).T
)
npt.assert_array_equal(
np.array(list(count_table.ids(axis='sample'))),
np.array(['sample1', 'sample2', 'sample3'])
)
npt.assert_array_equal(
np.array(list(count_table.ids(axis='observation'))),
np.array(['seq1', 'seq2', 'seq3', 'seq4', 'seq5', 'seq6']),
)
pdt.assert_frame_equal(known_map, mapping)
pdt.assert_frame_equal(known_summary, summary.to_dataframe())