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 main():
args = parse_args(sys.argv[1:])
if args.a is not None and args.t is not None:
sys.exit("Usage: -a PATH/ACC -t PATH/TAX -out PATH [-m JSON]")
experiments = []
if args.a is not None:
# Extract accession related experiments
if os.path.exists(args.a):
accfile = args.a
elif args.a[:3] in acctypes:
accfile = "tmp.acc"
with open(accfile, "w") as f:
f.write("\n".join(args.a.split(",")))
experiments.extend(SetupParallelDownload(accfile))
if args.t is not None:
# Extract tax id related experiments
if os.path.exists(args.t):
taxfile = args.a
else:
taxfile = "tmp.tax"
with open(taxfile, "w") as f:
f.write("\n".join(args.t.split(",")))
experiments.extend(SetupParallelDownload(taxfile))
# Remove doublicate experiments
experiments = list(set(experiments))
elen = len(experiments)
print("Found %s unique experiment Accessions IDs!" % (elen))
if elen > 0:
# Create out directory
cwd = os.getcwd()
out_dir = "%s/%s/" % (cwd, args.out)
if not os.path.exists(out_dir):
os.mkdir(out_dir)
os.chdir(out_dir)
# Split experiments in batches
epb = ceil(elen / float(args.nodes))
batches = [experiments[s : s + epb] for s in xrange(0, elen, epb)]
# Run batch downloads
ps = []
for batch_dir, eids in enumerate(batches):
# Save experiment IDs to file
batch_acc_list = "%s/%s.acc" % (out_dir, batch_dir)
with open(batch_acc_list, "w") as f:
f.write("\n".join(eids))
# Prepare cmdline
nargs = ["-a", batch_acc_list, "-out", str(batch_dir)]
if args.preserve:
nargs.append("-p")
if args.m is not None:
nargs.extend(["-m", "%s/%s" % (cwd, args.m)])
if args.all_runs_as_samples:
nargs.append("--all_runs_as_samples")
cmd = GetCMD("download-accession-list", nargs)
# Execute batch download
ps.append(Popen(cmd, shell=True, executable="/bin/bash"))
# Wait for all batches to finish
esum = 0
for p in ps:
esum += p.wait()
if esum == 0:
print("All batches finished succesfully!")
else:
print("Something failed!")
else:
print("No experiments could be found!")