def test_hmmscan_task_multithreaded(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('20aa-alitest.fa', td) as prot, \
TestData('20aa.hmm', td) as hmm, \
TemporaryFile(td) as out_single,\
TemporaryFile(td) as out_multi:
for n_threads in (2, 3, 4, 5):
db_task = tasks.get_hmmpress_task(
hmm, self.hmmpress_cfg)
aln_task_single = tasks.get_hmmscan_task(
prot, out_single, hmm, 1.0, 1, self.hmmscan_cfg)
aln_task_multi = tasks.get_hmmscan_task(
prot, out_multi, hmm, 1.0, n_threads,
self.hmmscan_cfg)
run_tasks([db_task, aln_task_single], ['run'])
run_task(aln_task_multi)
print(os.listdir(td), file=sys.stderr)
print(open(out_single).read())
alns_single = pd.concat(hmmscan_to_df_iter(out_single))
alns_multi = pd.concat(hmmscan_to_df_iter(out_multi))
self.assertTrue(all(alns_single['domain_i_evalue'].sort_values() == \
alns_multi['domain_i_evalue'].sort_values()))
def test_cmscan_task_multithreaded(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('rnaseP-bsu.fa', td) as transcript, \
TestData('rnaseP-eubact.c.cm', td) as cm, \
TemporaryFile(td) as out_single,\
TemporaryFile(td) as out_multi:
for n_threads in (2, 3, 4, 5):
db_task = tasks.get_cmpress_task(cm, self.cmpress_cfg)
aln_task_single = tasks.get_cmscan_task(
transcript, out_single, cm, 1.0, 1,
self.cmscan_cfg)
aln_task_multi = tasks.get_cmscan_task(
transcript, out_multi, cm, 1.0, n_threads,
self.cmscan_cfg)
run_tasks([db_task, aln_task_single], ['run'])
run_task(aln_task_multi)
alns_single = pd.concat(cmscan_to_df_iter(out_single))
alns_multi = pd.concat(cmscan_to_df_iter(out_multi))
self.assertTrue(all(alns_single['e_value'].sort_values() == \
alns_multi['e_value'].sort_values()))
def test_lastal_task_nucl_x_prot(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-protein.fa', td) as prot, \
TestData('test-transcript.fa', td) as tr, \
TemporaryFile(td) as out:
print(os.listdir(td), file=sys.stderr)
db_task = tasks.get_lastdb_task(prot, prot,
self.lastdb_cfg)
aln_task = tasks.get_lastal_task(tr,
prot,
out,
self.lastal_cfg,
translate=True,
cutoff=None)
run_tasks([db_task, aln_task], ['run'])
aln = ''.join(open(out).readlines())
print(aln, file=sys.stderr)
self.assertIn('SPAC212_RecQ_type_DNA_helicase_PROTEIN',
aln)
self.assertIn('SPAC212_RecQ_type_DNA_helicase_TRANSCRIPT',
aln)
self.assertIn('lambda',
aln,
msg='lambda missing, wrong LAST version?')
def test_save(self):
'''Test that FigureManager saves a plot.
'''
with TemporaryDirectory() as td:
with TemporaryFile(td) as filename:
with FigureManager(filename=filename + '.pdf') as (fig, ax):
ax.plot(range(10), range(10))
self.assertTrue(os.path.isfile(filename + '.pdf'))
def update_serialized_option_value(shell, option_table, option_name, expr):
"""
직렬화 된 값 안에서 문자열은 문자열의 길이를 기억하게 되어 있다. 문자열 치환이 직렬화된 값 안에서 일어났다면 문자열의 길이도 변경해야 한다.
:param shell:
:param option_table:
:param option_name:
:param expr:
:return:
"""
def repl(match):
return 's:%d:\"%s\";' % (len(match.group(2)), match.group(2))
query = '"SELECT option_value AS \'\' FROM %s WHERE option_name=\'%s\'"' % (
option_table, option_name)
cmd = shell.cmd_mysql(mysql_options='-s -r -e {}'.format(query))
m = local(cmd, capture=True)
expr = 's:(\d+):"(%s)";' % expr
replaced = escape_string(re.sub(expr, repl, m.stdout)).decode('utf-8')
f = TemporaryFile()
query = 'UPDATE %s SET option_value=\'%s\' WHERE option_name=\'%s\';' % (
option_table, replaced, option_name)
f.write(query)
f.close()
local(shell.cmd_mysql() + ' < ' + f.name)
def test_lastal_task_multithreaded(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-protein.fa', td) as prot, \
TestData('pom.50.fa', td) as tr, \
TemporaryFile(td) as out_single,\
TemporaryFile(td) as out_multi:
for n_threads in (3, 4, 5):
print(os.listdir(td), file=sys.stderr)
db_task = tasks.get_lastdb_task(
prot, prot, self.lastdb_cfg)
aln_task_single = tasks.get_lastal_task(
tr,
prot,
out_single,
self.lastal_cfg,
translate=True,
cutoff=None)
aln_task_multi = tasks.get_lastal_task(
tr,
prot,
out_multi,
self.lastal_cfg,
translate=True,
cutoff=None,
n_threads=n_threads)
run_tasks([db_task, aln_task_single, aln_task_multi],
['run'])
alns_single = MafParser(out_single).read()
alns_multi = MafParser(out_multi).read()
self.assertTrue(all(alns_single['E'].sort_values() == \
alns_multi['E'].sort_values()))
def test_exception(self):
'''Test that FigureManager handles internal exceptions properly.
'''
with TemporaryDirectory() as td:
with TemporaryFile(td) as filename:
# now we test the manager
try:
rregexp = self.assertRaisesRegex
except AttributeError:
rregexp = self.assertRaisesRegexp
with rregexp(RuntimeError, 'TEST'):
with FigureManager(filename=filename) as (fig, ax):
raise RuntimeError('TEST')
self.assertFalse(os.path.isfile(filename),
msg='Should not have saved a file.')
def test_hmmscan_task(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-protein.fa', td) as prot, \
TestData('test-profile.hmm', td) as hmm, \
TemporaryFile(td) as out:
db_task = tasks.get_hmmpress_task(hmm, self.hmmpress_cfg)
aln_task = tasks.get_hmmscan_task(prot, out, hmm, 1.0, 1,
self.hmmscan_cfg)
run_tasks([db_task, aln_task], ['run'])
print(os.listdir(td), file=sys.stderr)
aln = open(out).read()
print(aln)
self.assertEquals(aln.count('accession'), 2)
self.assertIn('i-Evalue', aln)
def test_lastal_task_uptodate(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-protein.fa', td) as prot, \
TemporaryFile(td) as out:
print(os.listdir(td), file=sys.stderr)
db_task = tasks.get_lastdb_task(prot, prot, self.lastdb_cfg)
aln_task = tasks.get_lastal_task(prot, prot, out, False,
None, 1, self.lastal_cfg)
# Run it once
run_tasks([db_task, aln_task], ['run'])
# Now run again and check the status
run_tasks([aln_task], ['run'])
status = check_status(aln_task)
self.assertEquals(status.status, 'up-to-date')
def test_cmscan_task(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-transcript.fa', td) as transcript, \
TestData('test-covariance-model.cm', td) as cm, \
TemporaryFile(td) as out:
db_task = tasks.get_cmpress_task(cm, self.cmpress_cfg)
aln_task = tasks.get_cmscan_task(transcript, out, cm, 1.0, 1,
self.cmscan_cfg)
run_tasks([db_task, aln_task], ['run'])
print(os.listdir(td), file=sys.stderr)
aln = ''.join(open(out).readlines())
print(aln, file=sys.stderr)
# TODO: better correctness check
self.assertEquals(aln.count('accession'), 2)
self.assertIn('E-value', aln)
def update_url(shell, replacement_items, old_string, new_string):
"""
데이터베이스 안의 내용을 치환.
"""
if old_string == new_string:
return
f = TemporaryFile()
for item in replacement_items:
query = "UPDATE `%s` SET `%s` = REPLACE(`%s`, \'%s\', \'%s\');\n" % (
item['table'], item['field'], item['field'], old_string,
new_string)
f.write(query)
f.close()
local(shell.cmd_mysql() + ' < ' + f.name)
def test_lastal_task_prot_x_prot(self):
with TemporaryDirectory() as td:
with Move(td):
with TestData('test-protein.fa', td) as prot, \
TemporaryFile(td) as out:
print(os.listdir(td), file=sys.stderr)
db_task = tasks.get_lastdb_task(prot, prot, self.lastdb_cfg)
aln_task = tasks.get_lastal_task(prot, prot, out, False,
None, 1, self.lastal_cfg)
run_tasks([db_task, aln_task], ['run'])
aln = ''.join(open(out).readlines())
print(aln, file=sys.stderr)
self.assertEquals(
aln.count('SPAC212_RecQ_type_DNA_helicase_PROTEIN'),
2)
self.assertIn('EG2=0', aln)
self.assertIn('E=0', aln)
self.assertIn('lambda', aln,
msg='lambda missing, wrong LAST version?')