def test_fasta_from_sqlite_same_clusters_different_rep_seq(self):
# same as `test_fasta_from_sqlite`, but invert the selected rep seq
# in each cluster by swapping the feature counts in the test data.
conn = sqlite3.connect(':memory:')
c = conn.cursor()
c.execute('CREATE TABLE feature_cluster_map'
'(feature_id TEXT PRIMARY KEY,cluster_id TEXT NOT NULL, '
'count INTEGER);')
s = [('feature1', 'r1', 15), ('feature2', 'r2', 24),
('feature3', 'r1', 204), ('feature4', 'r2', 4)]
c.executemany('INSERT INTO feature_cluster_map VALUES (?, ?, ?);', s)
conn.commit()
with tempfile.NamedTemporaryFile() as output_sequences_f:
_fasta_from_sqlite(conn, self.input_sequences_fp,
output_sequences_f.name)
obs_seqs = _read_seqs(output_sequences_f.name)
rep_seqs = _read_seqs(self.input_sequences)
exp_seqs = [
rep_seqs[1], # feature2
rep_seqs[2]
] # feature3
_relabel_seqs(exp_seqs, ['r2', 'r1'])
self.assertEqual(obs_seqs, exp_seqs)
def test_clusters_with_multiple_features_with_same_count(self):
# feature1 and feature3 cluster into r1, feature2 and feature4 cluster
# into r2. The features within a cluster have the same count, so this
# test should ensure that the right rep seq is picked for each cluster.
# The query in _fasta_from_sqlite should break ties by using the
# first feature when sorting the tied features alphabetically by id.
conn = sqlite3.connect(':memory:')
c = conn.cursor()
c.execute('CREATE TABLE feature_cluster_map'
'(feature_id TEXT PRIMARY KEY,cluster_id TEXT NOT NULL, '
'count INTEGER);')
s = [('feature1', 'r1', 204), ('feature2', 'r2', 4),
('feature3', 'r1', 204), ('feature4', 'r2', 4)]
c.executemany('INSERT INTO feature_cluster_map VALUES (?, ?, ?);', s)
conn.commit()
with tempfile.NamedTemporaryFile() as output_sequences_f:
_fasta_from_sqlite(conn, self.input_sequences_fp,
output_sequences_f.name)
obs_seqs = _read_seqs(output_sequences_f.name)
rep_seqs = _read_seqs(self.input_sequences)
exp_seqs = [rep_seqs[0], # feature1
rep_seqs[1]] # feature2
_relabel_seqs(exp_seqs, ['r1', 'r2'])
self.assertEqual(obs_seqs, exp_seqs)
def test_fasta_from_sqlite(self):
# artificially clustering feature1 and feature3 into r1, and
# feature2 and feature4 into r2.
conn = sqlite3.connect(':memory:')
c = conn.cursor()
c.execute('CREATE TABLE feature_cluster_map'
'(feature_id TEXT PRIMARY KEY,cluster_id TEXT NOT NULL, '
'count INTEGER);')
s = [('feature1', 'r1', 204), ('feature2', 'r2', 4),
('feature3', 'r1', 15), ('feature4', 'r2', 24)]
c.executemany('INSERT INTO feature_cluster_map VALUES (?, ?, ?);', s)
conn.commit()
with tempfile.NamedTemporaryFile() as output_sequences_f:
_fasta_from_sqlite(conn, self.input_sequences_fp,
output_sequences_f.name)
obs_seqs = _read_seqs(output_sequences_f.name)
rep_seqs = _read_seqs(self.input_sequences)
exp_seqs = [
rep_seqs[0], # feature1
rep_seqs[3]
] # feature4
_relabel_seqs(exp_seqs, ['r1', 'r2'])
self.assertEqual(obs_seqs, exp_seqs)