def download_fastq(self):
'''
Download Fastq associated with Accession from ENA
:param run_accession: Run Accession ID from ENA
:return: True
'''
try:
Path(self.dir).makedirs_p()
retcode = call(self.download, stdout=PIPE)
except OSError as e:
_logger.error('FastQ Failed: %s [%s]', self.accession, e)
_logger.error('CMD: %s', self.download)
Sequence.__errors[self.accession] = 'FastQ Failed'
self.error = True
else:
if retcode < 0:
_logger.error('Child was terminated by signal')
self.error = True
Sequence.__errors[self.accession] = 'Child was'\
'terminated'\
'(signal)'
else:
_logger.info('Success: %s', self.accession)
self.files = [
f.abspath() for f in Path(self.dir).files()
]
Sequence.__sequence_id += 1
def ProcessExperimentSeparate(experiment_id, json, batch_dir, sample_dir_id, preserve, failed_accession, skip_files=False):
m = ExtractExperimentMetadata(experiment_id, json)
if m.valid_metadata():
# Check if a run ID was submitted, and if so only process that
if experiment_id in m.runIDs: m.runIDs = [experiment_id]
# Process the runIDs as samples
_logger.info("Found Following Runs: %s", ', '.join(m.runIDs))
for runid in m.runIDs:
with TemporaryDirectory() as tmpdir:
os.chdir(batch_dir)
sample_dir = "%s/%s/"%(batch_dir, sample_dir_id)
if os.path.exists(sample_dir):
sfiles = [x for x in os.listdir(sample_dir) if any([y in x for y in ['fq','fastq']])]
else:
sfiles = []
if not preserve or not skip_files or len(sfiles) == 0:
sfiles = DownloadRunFiles(runid, tmpdir)
if sfiles is not None:
success = CreateSampleDir(sfiles, m, sample_dir, preserve, skip_files)
if success:
sample_dir_id += 1
else:
failed_accession.append(runid)
else:
_logger.error("Files could not be retrieved! (%s)", runid)
failed_accession.append(runid)
else:
_logger.error("Metadata Invalid! (%s) - %s", experiment_id, m.metadata.items())
failed_accession.append(experiment_id)
return sample_dir_id
def CreateSampleDir(sfiles, m, sample_dir, preserve=False, skip_files=False):
sample_dir = str(sample_dir)
if not skip_files and len(sfiles) == 0:
_logger.error("Error: No files were found! (%s)", sample_dir)
return False
if not os.path.exists(sample_dir):
_logger.info("Create sample dir: %s", sample_dir)
# Create 'sample' dir
os.mkdir(sample_dir)
# Move files from tmpdir to sample dir
for sf in sfiles: move(sf, sample_dir)
elif not preserve and not skip_files:
# Empty sample directory
for fn in os.listdir(sample_dir):
os.unlink("%s/%s"%(sample_dir, fn))
# Move files from tmpdir to sample dir
for sf in sfiles: move(sf, sample_dir)
# Update and create metadata file
try:
m.metadata["file_names"] = ' '.join(
[os.path.basename(sf).replace(' ','_')
for sf in sfiles
if not os.path.basename(sf) == 'meta.json']
)
m.save_metadata(sample_dir)
except ValueError, e:
_logger.error(e)
return False
def download_fastq_from_list(accession_list, output, json, preserve=False, all_runs_as_samples=False, skip_files=False):
"""
Get Fastq from list of IDs
:param accession_list: List of accessions
:param dir: Output folder
"""
metadata = []
cwd = os.getcwd()
with open(accession_list, 'r') as f:
# Setup batch dir
batch_dir = "%s/%s/"%(cwd, output)
if not os.path.exists(batch_dir): os.mkdir(batch_dir)
os.chdir(batch_dir)
# Set logging
_logger.Set(filename="%s/download-acceession-list.log"%batch_dir)
# Count samples in accession_list
n_samples = sum(1 for l in f)
f.seek(0)
_logger.info("Number of samples to download: %s", n_samples)
# Start progress bar
pbar = ProgressBar(
widgets = [ETA(), ' - ', Percentage(), ' : ', Bar()],
maxval = n_samples
).start()
pbar.update(0)
failed_accession = []
sample_dir_id = 0
for i, l in enumerate(f):
accession = l.strip()
if accession == '': continue
# Determine accession type
if accession[:3] in acctypes:
accession_type = acctypes[accession[:3]]
else:
_logger.error("unknown accession type for '%s'!", accession)
failed_accession.append(accession)
continue
_logger.info("Acc Found: %s (%s)", accession, accession_type)
if accession_type in ['study', 'sample']:
for experiment_id in ExtractExperimentIDs_acc(accession):
sample_dir_id = ProcessExperiment(
experiment_id, json, batch_dir,sample_dir_id, preserve,
failed_accession, all_runs_as_samples, skip_files)
elif accession_type == 'experiment':
sample_dir_id = ProcessExperiment(
accession, json, batch_dir,sample_dir_id, preserve,
failed_accession, all_runs_as_samples, skip_files)
elif accession_type == 'run':
sample_dir_id = ProcessExperiment(
accession, json, batch_dir,sample_dir_id, preserve,
failed_accession, all_runs_as_samples, skip_files)
pbar.update(i)
pbar.finish()
if failed_accession:
_logger.info("The following accessions were not downloaded!")
_logger.info('\n'.join(failed_accession))
else:
_logger.info("All accessions downloaded succesfully!")
def ProcessExperiment(experiment_id, json, batch_dir, sample_dir_id, preserve, failed_accession, all_runs_as_samples, skip_files=False):
_logger.info("Processing %s...", experiment_id)
if all_runs_as_samples:
sample_dir_id = ProcessExperimentSeparate(
experiment_id, json, batch_dir, sample_dir_id,
preserve, failed_accession, skip_files)
else:
sample_dir_id = ProcessExperimentCombined(
experiment_id, json, batch_dir, sample_dir_id,
preserve, failed_accession, skip_files)
return sample_dir_id
def DownloadRunFiles(runid, tmpdir):
# Download run files
try:
s = Sequence(runid, tmpdir)
s.download_fastq()
if not s.error:
_logger.info("Downloaded files: %s", ','.join(s.files))
return s.files
else: return None
except ValueError, e:
_logger.error(e)
return None
def ExtractData(self, query):
''' Extract Sample Metadata '''
new_platforms = []
new_seqtypes = []
# New approach using runinfo list
with openurl(self.sra_url1%(query)) as u:
headers = u.readline().split(',')
indexes = [(x, headers.index(x)) for x in ["Run", "Experiment",
"Sample", "SRAStudy", "BioSample", "Platform", "LibraryLayout",
"SampleName", "ScientificName", "CenterName"]]
for l in u:
l = l.strip()
if l == '': continue
if l[0] == '#': continue
d = l.split(',')
self.accessions['run'] = d[indexes[0][1]]
self.accessions['experiment'] = d[indexes[1][1]]
self.accessions['sample'] = d[indexes[2][1]]
self.accessions['study'] = d[indexes[3][1]]
self.accessions['biosample'] = d[indexes[4][1]]
platform = d[indexes[5][1]].lower()
if platform in platforms:
self['sequencing_platform'] = platforms[platform]
else:
self['sequencing_platform'] = 'unknown'
if not platform in new_platforms:
new_platforms.append(platform)
seqtype = d[indexes[6][1]].lower()
if seqtype in sequencing_types:
self['sequencing_type'] = sequencing_types[seqtype]
else:
self['sequencing_type'] = 'unknown'
if not seqtype in new_seqtypes:
new_seqtypes.append(seqtype)
self['sample_name'] = d[indexes[7][1]]
self['organism'] = d[indexes[8][1]]
self['collected_by'] = d[indexes[9][1]]
self['biosample'] = self.accessions['biosample']
break # Just use the first entry!
# Should be fixed to handle different query sequences!!!
with openurl(self.sra_url2%(query)) as u: qdata = u.read()
# Extract sample attributes
match = re.findall(r'Sample Attributes: (.+)\n', qdata)
lcs = {} # location parts
host = None
source = None
for answer in match:
for attributes in answer.split(';'):
stat = attributes.split('=')
att = stat[0].strip('/ ').lower().replace('\'', '')
val = stat[1].strip('\' ').replace('\'', '\`')
if att in ['geo_loc_name', 'geographic location']:
self.__interpret_loc(val)
elif att == 'serovar':
self['subtype']['serovar'] = val
elif att == 'mlst':
self['subtype']['mlst'] = val
elif att in ['scientific_name', 'scientific name']:
self['organism'] = val
elif att == 'strain':
self['strain'] = val
elif att in ['isolation_source', 'isolation source']:
source = val
elif att in ['host', 'specific_host', 'specific host']:
host = val
elif att == 'BioSample':
self['biosample'] = val
elif att in ['collection_date', 'collection date']:
self['collection_date'] = self.__format_date(
*self.__interpret_date(val)
)
if self['collection_date'] == '':
_logger.warning(
'Date Empty: %s, %s',
val, query
)
elif att in ['collected_by', 'collected by']:
self['collected_by'] = val
elif att in ['country', 'region', 'city', 'zip_code']:
lcs[att] = val
else:
self['notes'] = '%s %s: %s,' % (
self['notes'], att, val)
if lcs != {}:
h = ['country', 'region', 'city', 'zip_code']
self.__interpret_loc( ','.join([lcs[x] for x in h if x in lcs]))
# Handle Isolation source
cats = []
if host is not None:
for d in ontology:
cats = [d[k][0] for k in d.keys() if k in host.lower()]
if cats:
break
if not cats and host not in self.new_ontologies:
self.new_ontologies[host] = query
if (not cats or cats[0] == 'unknown') and source is not None:
for d in ontology:
cats = [d[k][0] for k in d.keys() if k in source.lower()]
if cats:
break
if not cats and source not in self.new_ontologies:
self.new_ontologies[source] = query
if cats:
self['isolation_source'] = cats[0]
_logger.warning(
'Source identified: %s (%s, %s), %s',
self['isolation_source'], host, source, query
)
else:
if host is None: host = 'unknown'
if source is None: source = 'unknown'
_logger.warning(
'Source not identified: (%s, %s), %s',
host, source, query
)
self['source_note'] = source
# Extract Run IDs associated with the sample
#Run #1: ERR276921, 1356661 spots, 271332200 bases
self.runIDs = re.findall(r'Run #\d+: (.+?),.+', qdata)
# Notify Curators By Email
_logger.info('Make mail? %s'%(mail is not None))
if mail is not None:
_logger.info('Any unknowns? %s'%(len(self.new_ontologies) > 0))
if len(self.new_ontologies) > 0:
_logger.debug(mail.test(
'New isolation source...',
'Sources not identified:\n%s\n'%(
'\n'.join(map(', '.join, self.new_ontologies.items()))
)
))
mail.send(
'New isolation source...',
'Sources not identified:\n%s\n'%(
'\n'.join(map(', '.join, self.new_ontologies.items()))
)
)
if len(new_platforms) > 0:
_logger.debug(mail.test(
'New platforms...',
'Platforms not accounted for:\n%s\n'%(
'\n'.join(new_platforms)
)
))
mail.send(
'New platforms...',
'Platforms not accounted for:\n%s\n'%(
'\n'.join(new_platforms)
)
)
elif len(self.new_ontologies) > 0:
_logger.debug(
"NO MAIL!",
'Sources not identified:\n%s\n'%(
'\n'.join(map(', '.join, self.new_ontologies.items()))
)
)
def ProcessExperimentCombined(experiment_id, json, batch_dir, sample_dir_id, preserve, failed_accession, skip_files=False):
m = ExtractExperimentMetadata(experiment_id, json)
if m.valid_metadata():
# Check if a run ID was submitted, and if so only process that
if experiment_id in m.runIDs: m.runIDs = [experiment_id]
# Process the runs as one sample
_logger.info("Found Following Runs: %s", ', '.join(m.runIDs))
with TemporaryDirectory() as tmpdir:
os.chdir(batch_dir)
sample_dir = "%s/%s/"%(batch_dir, sample_dir_id)
csfiles = []
if preserve and os.path.exists(sample_dir):
csfiles = [x for x in os.listdir(sample_dir) if any([y in x for y in ['fq','fastq']])]
if csfiles == [] and not skip_files:
sfiles = []
for runid in m.runIDs:
sf = DownloadRunFiles(runid, tmpdir)
if sf is not None:
sfiles.append(sf)
else:
_logger.error("Run files could not be retrieved! (%s)",
runid)
_logger.info("Found Following files sets:\n%s\n",
'\n'.join([', '.join(sf) for sf in sfiles]))
# Combine sfiles into one entry
if len(sfiles) > 1:
for file_no, file_set in enumerate(zip(*sfiles)):
ext = '.'.join(file_set[0].split('/')[-1].split('.')[1:])
if len(sfiles[0]) > 1:
new_file = "%s_%s.combined.%s"%(experiment_id,file_no+1, ext)
else:
new_file = "%s.combined.%s"%(experiment_id, ext)
with open(new_file, 'w') as nf:
for fn in file_set:
with open(fn, 'rb') as f:
nf.write(f.read())
if os.path.exists(new_file):
csfiles.append(new_file)
else:
_logger.error("Combined file creation failed! (%s: %s)",
experiment_id, file_no)
break
elif isinstance(sfiles[0], list):
csfiles = sfiles[0]
if csfiles == []:
_logger.error("Files could not be combined! (%s)",
experiment_id)
failed_accession.append(experiment_id)
if csfiles != [] or skip_files:
success = CreateSampleDir(csfiles, m, sample_dir, preserve, skip_files)
if success:
sample_dir_id += 1
else:
failed_accession.append(experiment_id)
else:
_logger.error("Files could not be retrieved! (%s)",
experiment_id)
failed_accession.append(experiment_id)
else:
_logger.error("Metadata Invalid! (%s) - %s", experiment_id, m.metadata.items())
failed_accession.append(experiment_id)
return sample_dir_id
