def test_override(self):
with open(self.get_data_path('expected/paired-override.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-override.fasta'),
'fasta',
constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
exp_md = qiime2.Metadata.load(
self.get_data_path('expected/paired-override-stats.tsv'))
# NOTE: the test data isn't interesting enough to be impacted by
# chimera_method or min_fold_parent_over_abundance.
table, rep_seqs, md = denoise_paired(
self.demux_seqs,
150,
150,
trim_left_f=10,
trim_left_r=10,
max_ee=20.5,
trunc_q=0,
n_threads=1,
n_reads_learn=2,
hashed_feature_ids=False,
chimera_method='consensus',
min_fold_parent_over_abundance=1.1)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs), _sort_seqs(exp_rep_seqs))
self.assertEqual(md, exp_md)
def test_trim_left_bigger_than_trunc_len(self):
with self.assertRaisesRegex(ValueError, 'trim_left_f'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_f=150)
with self.assertRaisesRegex(ValueError, 'trim_left_r'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_r=150)
# Shouldn't fail when `trunc_len_f=0`
denoise_paired(self.demux_seqs, 0, 150, trim_left_f=10)
# Shouldn't fail when `trunc_len_r=0`
denoise_paired(self.demux_seqs, 150, 0, trim_left_r=10)
def test_trim_left_bigger_than_trunc_len(self):
with self.assertRaisesRegex(ValueError, 'trim_left_f'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_f=150)
with self.assertRaisesRegex(ValueError, 'trim_left_r'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_r=150)
# Shouldn't fail when `trunc_len_f=0`
denoise_paired(self.demux_seqs, 0, 150, trim_left_f=10)
# Shouldn't fail when `trunc_len_r=0`
denoise_paired(self.demux_seqs, 150, 0, trim_left_r=10)
def test_defaults(self):
with open(self.get_data_path('expected/paired-default.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-default.fasta'),
'fasta',
constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
table, rep_seqs = denoise_paired(self.demux_seqs, 150, 150)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs), _sort_seqs(exp_rep_seqs))
def test_all_reads_filtered(self):
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 10000, 10000)
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 150, 10000)
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 10000, 150)
def test_all_reads_filtered(self):
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 10000, 10000)
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 150, 10000)
with self.assertRaisesRegex(ValueError, 'filter'):
denoise_paired(self.demux_seqs, 10000, 150)
def test_defaults(self):
with open(self.get_data_path('expected/paired-default.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-default.fasta'),
'fasta',
constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
# NOTE: changing the chimera_method parameter doesn't impact the
# results for this dataset
table, rep_seqs = denoise_paired(self.demux_seqs, 150, 150)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs), _sort_seqs(exp_rep_seqs))
def test_defaults(self):
with open(self.get_data_path('expected/paired-default.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-default.fasta'),
'fasta', constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
# NOTE: changing the chimera_method parameter doesn't impact the
# results for this dataset
table, rep_seqs = denoise_paired(self.demux_seqs, 150, 150)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs),
_sort_seqs(exp_rep_seqs))
def test_override(self):
with open(self.get_data_path('expected/paired-override.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-override.fasta'),
'fasta', constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
table, rep_seqs = denoise_paired(
self.demux_seqs, 150, 150, trim_left_f=10, trim_left_r=10,
max_ee=20.5, trunc_q=0, n_threads=0, n_reads_learn=2,
hashed_feature_ids=False)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs),
_sort_seqs(exp_rep_seqs))
def test_bad_values_fail(self):
# Just confirm that the machinery works, anything more specific is just
# restating the _valid_inputs dict which is more declarative than a
# unit-test anyways.
with self.assertRaisesRegex(ValueError, 'trunc_len_f'):
denoise_paired(self.demux_seqs, -1, 150)
with self.assertRaisesRegex(ValueError, 'trunc_len_r'):
denoise_paired(self.demux_seqs, 150, -1)
with self.assertRaisesRegex(ValueError, 'n_reads_learn'):
denoise_paired(self.demux_seqs, 150, 150, n_reads_learn=0)
def test_no_chimera_method(self):
with open(self.get_data_path('expected/paired-default.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-default.fasta'),
'fasta', constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
exp_md = qiime2.Metadata.load(
self.get_data_path('expected/paired-default-stats.tsv'))
table, rep_seqs, md = denoise_paired(self.demux_seqs, 150, 150,
chimera_method='none')
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs),
_sort_seqs(exp_rep_seqs))
self.assertEqual(md, exp_md)
def test_remove_empty(self):
with open(self.get_data_path('expected/paired-remove-empty-default.tsv'
)) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-default.fasta'),
'fasta', constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
exp_md = qiime2.Metadata.load(
self.get_data_path('expected/paired-default-stats.tsv'))
# NOTE: changing the chimera_method parameter doesn't impact the
# results for this dataset
table, rep_seqs, md = denoise_paired(self.demux_seqs, 150, 150,
retain_all_samples=False)
self.assertEqual(_sort_table(table), _sort_table(exp_table))
self.assertEqual(_sort_seqs(rep_seqs),
_sort_seqs(exp_rep_seqs))
self.assertEqual(md, exp_md)
def test_override(self):
with open(self.get_data_path('expected/paired-override.tsv')) as fh:
exp_table = biom.Table.from_tsv(fh, None, None, lambda x: x)
exp_rep_seqs = list(
skbio.io.read(self.get_data_path('expected/paired-override.fasta'),
'fasta', constructor=skbio.DNA))
for seq in exp_rep_seqs:
del seq.metadata['description']
# NOTE: the test data isn't interesting enough to be impacted by
# chimera_method or min_fold_parent_over_abundance.
table, rep_seqs = denoise_paired(
self.demux_seqs, 150, 150, trim_left_f=10, trim_left_r=10,
max_ee=20.5, trunc_q=0, n_threads=0, n_reads_learn=2,
hashed_feature_ids=False, chimera_method='consensus',
min_fold_parent_over_abundance=1.1)
self.assertEqual(table, exp_table)
self.assertEqual(_sort_seqs(rep_seqs),
_sort_seqs(exp_rep_seqs))
def test_bad_values_fail(self):
# Just confirm that the machinery works, anything more specific is just
# restating the _valid_inputs dict which is more declarative than a
# unit-test anyways.
with self.assertRaisesRegex(ValueError, 'trunc_len_f'):
denoise_paired(self.demux_seqs, -1, 150)
with self.assertRaisesRegex(ValueError, 'trunc_len_r'):
denoise_paired(self.demux_seqs, 150, -1)
with self.assertRaisesRegex(ValueError, 'n_reads_learn'):
denoise_paired(self.demux_seqs, 150, 150, n_reads_learn=0)
with self.assertRaisesRegex(ValueError, 'consensus'):
denoise_single(self.demux_seqs, 150, 150, chimera_method='foo')
def test_trim_left_bigger_than_trunc_len(self):
with self.assertRaisesRegex(ValueError, 'trim_left_f'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_f=150)
with self.assertRaisesRegex(ValueError, 'trim_left_r'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_r=150)
def test_trim_left_bigger_than_trunc_len(self):
with self.assertRaisesRegex(ValueError, 'trim_left_f'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_f=150)
with self.assertRaisesRegex(ValueError, 'trim_left_r'):
denoise_paired(self.demux_seqs, 150, 150, trim_left_r=150)
