def query_gsm(gsm, out_file, config = {}):
gsm = gsm[0]
out_dir = os.path.dirname(os.path.abspath(out_file))
name = utils.splitext_plus(os.path.basename(out_file))[0]
url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=sra\&term={0}\&retmode=json".format(gsm)
cmd = "curl {0}".format(url)
process = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
out = process.stdout.read()
data = json.loads(out)
ids = data.get("esearchresult", {}).get("idlist", [])
logger.debug("Get id sample for %s" % gsm)
if ids:
gsm_info = _query_info("sra", ids[-1])
print(gsm_info)
srrall = []
for srr in gsm_info:
srrall.append(_create_link(srr))
logger.debug("Get FTP link for %s : %s" % (ids[-1], srrall))
outs = []
for srx in srrall:
sra_dir = utils.safe_makedir(os.path.join(out_dir, name))
srafiles = _download_srx(gsm, srx, sra_dir)
logger.debug("Get SRA for %s: %s" % (gsm, " ".join(srafiles)))
if srafiles:
for sra in srafiles:
outs.extend(_convert_fastq(sra, out_dir))
logger.debug("Get FASTQ for %s: %s" % (gsm, " ".join(outs)))
if outs:
files = combine_pairs(outs)
out_file = fastq.merge(files, out_file, config)
return out_file
def query_gsm(gsm, out_file, config = {}):
gsm = gsm[0]
out_dir = os.path.dirname(out_file)
url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=gds\&term={0}\&retmode=json".format(gsm)
cmd = "curl {0}".format(url)
process = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
out = process.stdout.read()
data = json.loads(out)
ids = data.get("esearchresult", {}).get("idlist", [])
logger.debug("Get id sample for %s" % gsm)
if ids:
gsm_info = _query_info("gds", ids[-1])
srxlist = gsm_info.get("result", {}).get(ids[-1], {}).get("extrelations", {})
srxall = []
for srxe in srxlist:
if srxe.get("targetftplink", None):
srxall.append(srxe["targetftplink"])
logger.debug("Get FTP link for %s : %s" % (ids[-1], srxall))
outs = []
for srx in srxall:
srafiles = _download_srx(gsm, srx, out_dir)
logger.debug("Get SRA for %s: %s" % (gsm, " ".join(srafiles)))
if srafiles:
for sra in srafiles:
outs.extend(_convert_fastq(sra, out_dir))
logger.debug("Get FASTQ for %s: %s" % (gsm, " ".join(outs)))
if outs:
files = combine_pairs(outs)
out_file = fastq.merge(files, out_file, config)
return out_file
def query_gsm(gsm, out_file, config = {}):
gsm = gsm[0]
out_dir = os.path.dirname(os.path.abspath(out_file))
name = utils.splitext_plus(os.path.basename(out_file))[0]
url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?db=sra\&term={0}\&retmode=json".format(gsm)
cmd = "curl {0}".format(url)
process = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, shell=True)
out = process.stdout.read()
data = json.loads(out)
ids = data.get("esearchresult", {}).get("idlist", [])
logger.debug("Get id sample for %s" % gsm)
if ids:
gsm_info = _query_info("sra", ids[-1])
logger.debug("gsm_info:%s" % gsm_info)
srrall = []
for srr in gsm_info:
srrall.append(_create_link(srr))
logger.debug("Get FTP link for %s : %s" % (ids[-1], srrall))
outs = []
for srx in srrall:
sra_dir = utils.safe_makedir(os.path.join(out_dir, name))
srafiles = _download_srx(srx, sra_dir)
if srafiles:
logger.debug("Get SRA for %s: %s" % (gsm, " ".join(srafiles)))
for sra in srafiles:
fastq_fn = _convert_fastq(sra, out_dir)
if fastq_fn:
outs.extend(fastq_fn)
logger.debug("Get FASTQ for %s: %s" % (gsm, " ".join(outs)))
if outs:
files = combine_pairs(outs)
out_file = fastq.merge(files, out_file, config)
return out_file
def query_srr(sra, out_file, config={}):
sra = sra[0]
outs = []
out_dir = os.path.dirname(os.path.abspath(out_file))
name = utils.splitext_plus(os.path.basename(out_file))[0]
srrall = []
for srr in sra:
srrall.append(_create_link(srr))
logger.debug("Get FTP link for %s : %s" % (name, srrall))
for srx in srrall:
sra_dir = utils.safe_makedir(os.path.join(out_dir, name))
srafiles = _download_srx(srx, sra_dir)
if srafiles:
logger.debug("Get SRA for %s: %s" % (sra, " ".join(srafiles)))
for sra in srafiles:
fastq_fn = _convert_fastq(sra, out_dir)
if fastq_fn:
outs.extend(fastq_fn)
logger.debug("Get FASTQ for %s: %s" % (sra, " ".join(outs)))
if outs:
files = combine_pairs(outs)
out_file = fastq.merge(files, out_file, config)
return out_file
def query_srr(sra, out_file, config = {}):
sra = sra[0]
outs = []
out_dir = os.path.dirname(os.path.abspath(out_file))
name = utils.splitext_plus(os.path.basename(out_file))[0]
srrall = []
for srr in sra:
srrall.append(_create_link(srr))
logger.debug("Get FTP link for %s : %s" % (name, srrall))
for srx in srrall:
sra_dir = utils.safe_makedir(os.path.join(out_dir, name))
srafiles = _download_srx(srx, sra_dir)
if srafiles:
logger.debug("Get SRA for %s: %s" % (sra, " ".join(srafiles)))
for sra in srafiles:
fastq_fn = _convert_fastq(sra, out_dir)
if fastq_fn:
outs.extend(fastq_fn)
logger.debug("Get FASTQ for %s: %s" % (sra, " ".join(outs)))
if outs:
files = combine_pairs(outs)
out_file = fastq.merge(files, out_file, config)
return out_file