def test_add_raw_data(self):
self._make_sandbox()
new = Study.create(
User('*****@*****.**'), 'NOT Identification of the '
'Microbiomes for Cannabis Soils', [1], self.info)
new.add_raw_data([RawData(1), RawData(2)])
obs = self.conn_handler.execute_fetchall(
"SELECT * FROM qiita.study_raw_data WHERE study_id=%s", (new.id, ))
self.assertEqual(obs, [[new.id, 1], [new.id, 2]])
def test_remove_filepath_errors(self):
fp = join(self.db_test_raw_dir, '1_s_G1_L001_sequences.fastq.gz')
with self.assertRaises(QiitaDBError):
RawData(1).remove_filepath(fp)
# filepath doesn't belong to that raw data
with self.assertRaises(ValueError):
RawData(2).remove_filepath(fp)
# the raw data has been linked to more than 1 study so it can't be
# unliked
Study(2).add_raw_data([RawData(2)])
with self.assertRaises(QiitaDBError):
RawData(2).remove_filepath(fp)
def test_add_raw_data_private(self):
new = Study.create(
User('*****@*****.**'), 'NOT Identification of the '
'Microbiomes for Cannabis Soils', [1], self.info)
new.status = 'private'
with self.assertRaises(QiitaDBStatusError):
new.add_raw_data([RawData(2)])
def test_get_preprocess_fastq_cmd(self):
raw_data = RawData(1)
params = PreprocessedIlluminaParams(1)
prep_template = PrepTemplate(1)
obs_cmd, obs_output_dir = _get_preprocess_fastq_cmd(
raw_data, prep_template, params)
get_raw_path = partial(join, self.db_dir, 'raw_data')
seqs_fp = get_raw_path('1_s_G1_L001_sequences.fastq.gz')
bc_fp = get_raw_path('1_s_G1_L001_sequences_barcodes.fastq.gz')
exp_cmd_1 = ("split_libraries_fastq.py --store_demultiplexed_fastq -i "
"{} -b {} "
"-m ".format(seqs_fp, bc_fp))
exp_cmd_2 = ("-o {0} --barcode_type golay_12 --max_bad_run_length 3 "
"--max_barcode_errors 1.5 "
"--min_per_read_length_fraction 0.75 "
"--phred_quality_threshold 3 "
"--sequence_max_n 0".format(obs_output_dir))
# We are splitting the command into two parts because there is no way
# that we can know the filepath of the mapping file. We thus split the
# command on the mapping file path and we check that the two parts
# of the commands is correct
obs_cmd_1 = obs_cmd[:len(exp_cmd_1)]
obs_cmd_2 = obs_cmd[len(exp_cmd_1):].split(" ", 1)[1]
self.assertEqual(obs_cmd_1, exp_cmd_1)
self.assertEqual(obs_cmd_2, exp_cmd_2)
def add_files_to_raw_data(raw_data_id, filepaths):
"""Add files to raw data
Needs to be dispachable because it moves large files
"""
rd = RawData(raw_data_id)
rd.add_filepaths(filepaths)
def test_delete(self):
# the raw data doesn't exist
with self.assertRaises(QiitaDBUnknownIDError):
RawData.delete(1000, 1)
# the raw data and the study id are not linked or
# the study doesn't exits
with self.assertRaises(QiitaDBError):
RawData.delete(1, 1000)
# the raw data has prep templates
with self.assertRaises(QiitaDBError):
RawData.delete(1, 1)
# the raw data has linked files
with self.assertRaises(QiitaDBError):
RawData.delete(3, 1)
# the raw data is linked to a study that has not prep templates
Study(2).add_raw_data([RawData(1)])
RawData.delete(1, 2)
# delete raw data
self.assertTrue(RawData.exists(2))
RawData.delete(2, 1)
self.assertFalse(RawData.exists(2))
def unlink_all_files(raw_data_id):
"""Removes all files from raw data
Needs to be dispachable because it does I/O and a lot of DB calls
"""
rd = RawData(raw_data_id)
rd.clear_filepaths()
def test_link_filepaths_status_setter(self):
rd = RawData(1)
self.assertEqual(rd.link_filepaths_status, 'idle')
rd._set_link_filepaths_status('linking')
self.assertEqual(rd.link_filepaths_status, 'linking')
rd._set_link_filepaths_status('unlinking')
self.assertEqual(rd.link_filepaths_status, 'unlinking')
rd._set_link_filepaths_status('failed: error')
self.assertEqual(rd.link_filepaths_status, 'failed: error')
def remove_add_prep_template(self, fp_rpt, raw_data_id, study,
data_type_id, investigation_type):
"""add prep templates"""
pt_id = PrepTemplate.create(load_template_to_dataframe(fp_rpt),
RawData(raw_data_id), study,
_to_int(data_type_id),
investigation_type=investigation_type).id
remove(fp_rpt)
return pt_id
def test_get_preprocess_fasta_cmd_sff(self):
raw_data = RawData(3)
params = Preprocessed454Params(1)
prep_template = PrepTemplate(1)
obs_cmd, obs_output_dir = _get_preprocess_fasta_cmd(
raw_data, prep_template, params)
get_raw_path = partial(join, self.db_dir, 'raw_data')
seqs_fp = [
get_raw_path('preprocess_test1.sff'),
get_raw_path('preprocess_test2.sff')
]
exp_cmd_1 = ' '.join(
["process_sff.py",
"-i %s" % seqs_fp[0],
"-o %s" % obs_output_dir])
exp_cmd_2 = ' '.join(
["process_sff.py",
"-i %s" % seqs_fp[1],
"-o %s" % obs_output_dir])
fasta_files = ','.join([
join(obs_output_dir, "preprocess_test1.fna"),
join(obs_output_dir, "preprocess_test2.fna")
])
qual_files = ','.join([
join(obs_output_dir, "preprocess_test1.qual"),
join(obs_output_dir, "preprocess_test2.qual")
])
exp_cmd_3a = ' '.join(["split_libraries.py", "-f %s" % fasta_files])
exp_cmd_3b = ' '.join([
"-q %s" % qual_files, "-d",
"-o %s" % obs_output_dir,
params.to_str()
])
exp_cmd_4 = ' '.join([
"convert_fastaqual_fastq.py",
"-f %s/seqs.fna" % obs_output_dir,
"-q %s/seqs_filtered.qual" % obs_output_dir,
"-o %s" % obs_output_dir, "-F"
])
obs_cmds = obs_cmd.split('; ')
# We are splitting the command into two parts because there is no way
# that we can know the filepath of the mapping file. We thus split the
# command on the mapping file path and we check that the two parts
# of the commands is correct
obs_cmd_3a, obs_cmd_3b_temp = obs_cmds[2].split(' -m ', 1)
obs_cmd_3b = obs_cmd_3b_temp.split(' ', 1)[1]
self.assertEqual(obs_cmds[0], exp_cmd_1)
self.assertEqual(obs_cmds[1], exp_cmd_2)
self.assertEqual(obs_cmd_3a, exp_cmd_3a)
self.assertEqual(obs_cmd_3b, exp_cmd_3b)
self.assertEqual(obs_cmds[3], exp_cmd_4)
def test_get_filepaths(self):
"""Correctly returns the filepaths to the raw files"""
rd = RawData(1)
obs = rd.get_filepaths()
exp = [(join(self.db_test_raw_dir,
'1_s_G1_L001_sequences.fastq.gz'), "raw_sequences"),
(join(self.db_test_raw_dir,
'1_s_G1_L001_sequences_barcodes.fastq.gz'),
"raw_barcodes")]
self.assertEqual(obs, exp)
def setUp(self):
metadata_dict = {
'SKB8.640193': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status_id': 1,
'data_type_id': 2,
'str_column': 'Value for sample 1'
},
'SKD8.640184': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status_id': 1,
'data_type_id': 2,
'str_column': 'Value for sample 2'
},
'SKB7.640196': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status_id': 1,
'data_type_id': 2,
'str_column': 'Value for sample 3'
}
}
self.metadata = pd.DataFrame.from_dict(metadata_dict, orient='index')
self.test_raw_data = RawData(1)
fd, seqs_fp = mkstemp(suffix='_seqs.fastq')
close(fd)
fd, barcodes_fp = mkstemp(suffix='_barcodes.fastq')
close(fd)
filepaths = [(seqs_fp, 1), (barcodes_fp, 2)]
with open(seqs_fp, "w") as f:
f.write("\n")
with open(barcodes_fp, "w") as f:
f.write("\n")
self.new_raw_data = RawData.create(2, filepaths, [Study(1)])
db_test_raw_dir = join(get_db_files_base_dir(), 'raw_data')
db_seqs_fp = join(db_test_raw_dir, "3_%s" % basename(seqs_fp))
db_barcodes_fp = join(db_test_raw_dir, "3_%s" % basename(barcodes_fp))
self._clean_up_files = [db_seqs_fp, db_barcodes_fp]
self.tester = PrepTemplate(1)
self.exp_sample_ids = {
'SKB1.640202', 'SKB2.640194', 'SKB3.640195', 'SKB4.640189',
'SKB5.640181', 'SKB6.640176', 'SKB7.640196', 'SKB8.640193',
'SKB9.640200', 'SKD1.640179', 'SKD2.640178', 'SKD3.640198',
'SKD4.640185', 'SKD5.640186', 'SKD6.640190', 'SKD7.640191',
'SKD8.640184', 'SKD9.640182', 'SKM1.640183', 'SKM2.640199',
'SKM3.640197', 'SKM4.640180', 'SKM5.640177', 'SKM6.640187',
'SKM7.640188', 'SKM8.640201', 'SKM9.640192'
}
def get_raw_data_from_other_studies(user, study):
"""Retrieves a tuple of raw_data_id and the last study title for that
raw_data
"""
d = {}
for sid in user.user_studies:
if sid == study.id:
continue
for rdid in Study(sid).raw_data():
d[int(rdid)] = Study(RawData(rdid).studies[-1]).title
return d
def create_raw_data(self, study, user, callback):
"""Adds a (new) raw data to the study
Parameters
----------
study : Study
The current study object
user : User
The current user object
callback : function
The callback function to call with the results once the processing
is done
"""
msg = "Raw data successfully added"
msg_level = "success"
# Get the arguments needed to create a raw data object
filetype = self.get_argument('filetype', None)
previous_raw_data = self.get_argument('previous_raw_data', None)
if filetype and previous_raw_data:
# The user selected a filetype and an existing raw data
msg = ("You can not specify both a new raw data and a previously "
"used one")
msg_level = "danger"
elif filetype:
# We are creating a new raw data object
try:
rd_id = RawData.create(filetype, [study]).id
except (TypeError, QiitaDBColumnError, QiitaDBExecutionError,
QiitaDBDuplicateError, IOError, ValueError, KeyError,
CParserError) as e:
msg = html_error_message % (
"creating a new raw data object for study:",
str(study.id), str(e))
msg_level = "danger"
elif previous_raw_data:
previous_raw_data = previous_raw_data.split(',')
raw_data = [RawData(rd) for rd in previous_raw_data]
study.add_raw_data(raw_data)
rd_id = raw_data[0].id
else:
# The user did not provide a filetype neither an existing raw data
# If using the interface, we should never reach this if, but
# better be safe than sorry
msg = ("You should choose a filetype for a new raw data or "
"choose a raw data previously used")
msg_level = "danger"
rd_id = None
callback((msg, msg_level, 'raw_data_tab', rd_id, None))
def render(self, prep, study_id, is_editable, ena_terms, study_status,
user_defined_terms):
# Check if the request came from a local source
is_local_request = self._is_local()
prep_id = prep.id
data_type = prep.data_type()
raw_data = RawData(prep.raw_data)
filepaths = prep.get_filepaths()
investigation_type = prep.investigation_type
preprocessed_data = prep.preprocessed_data
preprocessing_status = prep.preprocessing_status
if raw_data.filetype in ('SFF', 'FASTA'):
param_iter = Preprocessed454Params.iter()
elif raw_data.filetype == 'FASTQ':
param_iter = PreprocessedIlluminaParams.iter()
else:
raise ValueError("Don't know what to do but this exception will "
"never actually get shown anywhere because why "
"would you want to see tracebacks?")
preprocess_options = []
for param in param_iter:
text = ("<b>%s:</b> %s" % (k, v)
for k, v in viewitems(param.values))
preprocess_options.append(
(param.id, param.name, '<br>'.join(text)))
# Unfortunately, both the prep template and the qiime mapping files
# have the sample type. The way to differentiate them is if we have
# the substring 'qiime' in the basename
_fp_type = (lambda fp: "Qiime mapping"
if 'qiime' in basename(fp) else "Prep template")
filepaths = [(id_, fp, _fp_type(fp)) for id_, fp in filepaths]
return self.render_string(
"study_description_templates/prep_template_panel.html",
prep_id=prep_id,
data_type=data_type,
filepaths=filepaths,
investigation_type=investigation_type,
preprocessed_data=preprocessed_data,
preprocessing_status=preprocessing_status,
study_id=study_id,
is_local_request=is_local_request,
is_editable=is_editable,
ena_terms=ena_terms,
study_status=study_status,
user_defined_terms=user_defined_terms,
preprocess_options=preprocess_options)
def remove_add_study_template(self, raw_data, study_id, fp_rsp):
"""Replace prep templates, raw data, and sample template with a new one
"""
for rd in raw_data():
rd = RawData(rd)
for pt in rd.prep_templates:
if PrepTemplate.exists(pt):
PrepTemplate.delete(pt)
if SampleTemplate.exists(study_id):
SampleTemplate.delete(study_id)
SampleTemplate.create(load_template_to_dataframe(fp_rsp),
Study(study_id))
remove(fp_rsp)
def setUp(self):
self.raw_data = RawData(1)
self.study = Study(1)
self.params_table = "preprocessed_sequence_illumina_params"
self.params_id = 1
fd, self.fna_fp = mkstemp(suffix='_seqs.fna')
close(fd)
fd, self.qual_fp = mkstemp(suffix='_seqs.qual')
close(fd)
self.filepaths = [(self.fna_fp, 4), (self.qual_fp, 5)]
self.db_test_ppd_dir = join(get_db_files_base_dir(),
'preprocessed_data')
self.ebi_submission_accession = "EBI123456-A"
self.ebi_study_accession = "EBI123456-B"
with open(self.fna_fp, "w") as f:
f.write("\n")
with open(self.qual_fp, "w") as f:
f.write("\n")
self._clean_up_files = []
def post(self):
study_id = int(self.get_argument('study_id'))
prep_template_id = int(self.get_argument('prep_template_id'))
raw_data = RawData(PrepTemplate(prep_template_id).raw_data)
param_id = int(self.get_argument('preprocessing_parameters_id'))
# Get the preprocessing parameters
if raw_data.filetype == 'FASTQ':
param_constructor = PreprocessedIlluminaParams
elif raw_data.filetype in ('FASTA', 'SFF'):
param_constructor = Preprocessed454Params
else:
raise ValueError('Unknown filetype')
job_id = submit(self.current_user.id, preprocessor, study_id,
prep_template_id, param_id, param_constructor)
self.render('compute_wait.html',
job_id=job_id, title='Preprocessing',
completion_redirect='/study/description/%d?top_tab='
'raw_data_tab&sub_tab=%s&prep_tab=%s'
% (study_id, raw_data.id,
prep_template_id))
def get_raw_data(rdis):
"""Get all raw data objects from a list of raw_data_ids"""
return [RawData(rdi) for rdi in rdis]
def test_studies(self):
"""Correctly returns the study ids"""
rd = RawData(1)
self.assertEqual(rd.studies, [1])
def test_get_qiime_minimal_mapping_multiple(self):
# We need to create a prep template in which we have different run
# prefix values, so we can test this case
metadata_dict = {
'SKB8.640193': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status': 'EMP',
'str_column': 'Value for sample 1',
'linkerprimersequence': 'GTGCCAGCMGCCGCGGTAA',
'barcodesequence': 'GTCCGCAAGTTA',
'run_prefix': "s_G1_L001_sequences",
'platform': 'ILLUMINA',
'library_construction_protocol': 'AAA',
'experiment_design_description': 'BBB'
},
'SKD8.640184': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status': 'EMP',
'str_column': 'Value for sample 2',
'linkerprimersequence': 'GTGCCAGCMGCCGCGGTAA',
'barcodesequence': 'CGTAGAGCTCTC',
'run_prefix': "s_G1_L001_sequences",
'platform': 'ILLUMINA',
'library_construction_protocol': 'AAA',
'experiment_design_description': 'BBB'
},
'SKB7.640196': {
'center_name': 'ANL',
'center_project_name': 'Test Project',
'ebi_submission_accession': None,
'EMP_status': 'EMP',
'str_column': 'Value for sample 3',
'linkerprimersequence': 'GTGCCAGCMGCCGCGGTAA',
'barcodesequence': 'CCTCTGAGAGCT',
'run_prefix': "s_G1_L002_sequences",
'platform': 'ILLUMINA',
'library_construction_protocol': 'AAA',
'experiment_design_description': 'BBB'
}
}
md_template = pd.DataFrame.from_dict(metadata_dict, orient='index')
prep_template = PrepTemplate.create(md_template, RawData(2), Study(1),
'16S')
out_dir = mkdtemp()
obs_fps = sorted(_get_qiime_minimal_mapping(prep_template, out_dir))
exp_fps = sorted([
join(out_dir, 's_G1_L001_sequences_MMF.txt'),
join(out_dir, 's_G1_L002_sequences_MMF.txt')
])
# Check that the returned list is as expected
self.assertEqual(obs_fps, exp_fps)
# Check that the file exists
for fp in exp_fps:
self.assertTrue(exists(fp))
# Check the contents of the file
for fp, contents in zip(exp_fps, [EXP_PREP_1, EXP_PREP_2]):
with open(fp, "U") as f:
self.assertEqual(f.read(), contents)
def get_raw_data(self, rdis, callback):
"""Get all raw data objects from a list of raw_data_ids"""
callback([RawData(rdi) for rdi in rdis])
def test_data_type(self):
"""Correctly returns the data_type of raw_data"""
rd = RawData(1)
self.assertEqual(rd.data_type(), "18S")
def test_not_equal(self):
"""Not equals works with object of the same type"""
new = RawData(2)
self.assertNotEqual(self.tester, new)
def test_equal(self):
"""Equality works with two objects pointing to the same instance"""
new = RawData(1)
self.assertEqual(self.tester, new)
def test_init_error_inexistent(self):
"""Raises an error when instantiating an object that does not exists"""
with self.assertRaises(QiitaDBUnknownIDError):
RawData(10)
def setUp(self):
# We need an actual subclass in order to test the equality functions
self.tester = RawData(1)
def test_data_type_id(self):
"""Correctly returns the data_type of raw_data"""
rd = RawData(1)
self.assertEqual(rd.data_type(ret_id=True), 2)
def test_clear_filepaths_error(self):
with self.assertRaises(QiitaDBError):
RawData(1).clear_filepaths()
def _construct_job_graph(self, study, prep_template, params):
"""Constructs the workflow graph to preprocess a study
The steps performed to preprocess a study are:
1) Execute split libraries
2) Add the new preprocessed data to the DB
Parameters
----------
study : Study
The study to preprocess
prep_template : PrepTemplate
The prep template to use for the preprocessing
params : BaseParameters
The parameters to use for preprocessing
"""
self.prep_template = prep_template
self._logger = stderr
raw_data = RawData(prep_template.raw_data)
# Change the prep_template preprocessing_status t
self.prep_template.preprocessing_status = 'preprocessing'
# STEP 1: Preprocess the study
preprocess_node = "PREPROCESS"
# Check the raw data filetype to know which command generator we
# should use
filetype = raw_data.filetype
if filetype == "FASTQ":
cmd_generator = _get_preprocess_fastq_cmd
insert_preprocessed_data = _insert_preprocessed_data
elif filetype in ('FASTA', 'SFF'):
cmd_generator = _get_preprocess_fasta_cmd
insert_preprocessed_data = _insert_preprocessed_data
else:
raise NotImplementedError(
"Raw data %s cannot be preprocessed, filetype %s not supported"
% (raw_data.id, filetype))
# Generate the command
cmd, output_dir = cmd_generator(raw_data, self.prep_template, params)
self._job_graph.add_node(preprocess_node,
func=system_call,
args=(cmd, ),
job_name="Construct preprocess command",
requires_deps=False)
# This step is currently only for data types in which we need to store,
# demultiplexed sequences. Since it is the only supported data type at
# this point, it is ok the leave it here. However, as new data types
# become available, we will need to think a better way of doing this.
demux_node = "GEN_DEMUX_FILE"
self._job_graph.add_node(demux_node,
func=generate_demux_file,
args=(output_dir, ),
job_name="Generated demux file",
requires_deps=False)
self._job_graph.add_edge(preprocess_node, demux_node)
# STEP 2: Add preprocessed data to DB
insert_preprocessed_node = "INSERT_PREPROCESSED"
self._job_graph.add_node(insert_preprocessed_node,
func=insert_preprocessed_data,
args=(study, params, self.prep_template,
output_dir),
job_name="Store preprocessed data",
requires_deps=False)
self._job_graph.add_edge(demux_node, insert_preprocessed_node)
self._dirpaths_to_remove.append(output_dir)
