def ExtractExperimentIDs_tax(taxid):
''' Extract experiments which have runs associated from taxid
>>> ExtractExperimentIDs_tax('211968')
['SRX1308653', 'SRX1308716', 'SRX1308789', 'SRX1308879', 'SRX337751']
'''
ena_url = ('http://www.ebi.ac.uk/ena/data/warehouse/search?'
'query="tax_tree(%s)"&'
'result=read_experiment')%(taxid)
countquery = '&resultcount'
display = '&display=report&fields=experiment_accession'
# Find number of entries for the provided taxid
count = 0
with openurl(ena_url+countquery) as u:
for l in u:
l = l.strip()
if ':' in l:
tmp = l.split(':')
if tmp[0] == 'Number of results':
count = int(tmp[1].replace(',',''))
# Extract experiment IDs
experiments = []
if count > 0:
length = 100000
pages = ceil(count/float(length))
for p in xrange(pages):
page_offset = '&offset=%s&length=%s'%(p*length+1, length)
with openurl(ena_url+display+page_offset) as u:
header = u.readline()
for l in u:
l = l.strip()
if l[:3] in acctypes and acctypes[l[:3]] == 'experiment':
experiments.append(l)
else:
print("Unknown Experiment ID: %s (taxid=%s)"%(l,taxid))
return experiments
def ExtractTaxIDfromSearchTerm(query):
''' Extract taxonomy ID from NCBI taxonomy search
>>> ExtractTaxIDfromSearchTerm('Salmonella')
590
'''
ncbi_url = 'http://www.ncbi.nlm.nih.gov/taxonomy/?term=%s&report=taxid'%(
query)
# Find number of entries for the provided taxid
taxid = None
with openurl(ncbi_url) as u:
for l in u:
# remove html tags
l = re.sub('<.+?>', '', l)
l = l.strip()
if l == '': continue
try: taxid = int(l)
except: print("Error: Unhandled result from taxid search! (%s)"%l)
return taxid
def ExtractExperimentIDs_acc(sample_accession):
''' Extract experiments which have runs associated
>>> ExtractExperimentIDs_acc('ERS397989')
['ERX385098']
>>> ExtractExperimentIDs_acc('SRS331977')
['SRX146831', 'SRX365746', 'SRX146834', 'SRX146829', 'SRX146822', 'SRX146814', 'SRX146806']
'''
experiments = {}
sra_url = 'http://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?save=efetch&db=sra&rettype=runinfo&term=%s'
with openurl(sra_url%(sample_accession)) as u:
headers = u.readline()
try:
idx = headers.split(',').index("Experiment")
except:
print headers
else:
for l in u:
l = l.strip()
if l == '': continue
if l[0] == '#': continue
exp = l.split(',')[idx].strip()
if not exp in experiments: experiments[exp] = 1
return experiments.keys()
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()))
)
)
