def action_cp(args):
rg = nebula.warpdrive.RemoteGalaxy(args.url, args.api_key)
if not args.dir:
docstore = from_url(args.dst)
else:
if not os.path.exists(args.dst):
os.mkdir(args.dst)
for hda in rg.get_history_contents(args.src):
if hda['visible']:
if args.filter is None or re.search(args.filter, hda['name']):
if hda['name'] not in args.exclude:
print hda['name']
meta = rg.get_dataset(hda['id'], 'hda')
if args.dir:
dst_path = os.path.join(args.dst, hda['name'])
rg.download(meta['download_url'], dst_path)
else:
meta['id'] = meta['uuid'] #use the glocal id
hda = Target(uuid=meta['uuid'])
docstore.create(hda)
path = docstore.get_filename(hda)
rg.download(meta['download_url'], path)
docstore.update_from_file(hda)
def run_audit(args):
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
donor_map = {}
for id, ent in docstore.filter(state="ok"):
if ent['visible']:
if docstore.size(Target(id)) > 0:
donor = None
for i in ent['tags']:
t = i.split(":")
if t[0] == "donor":
donor = t[1]
if donor not in donor_map:
donor_map[donor] = {}
donor_map[donor][ent['name']] = id
for ent in synqueue.listAssignments(syn, list_all=True, **config):
if ent['meta']['Submitter_donor_ID'] in donor_map:
print ent['meta']['Submitter_donor_ID'], len(
donor_map[ent['meta']['Submitter_donor_ID']])
def run_upload(args):
# syn = synapseclient.Synapse()
# syn.login()
docstore = from_url(args.out_base)
donor_map = {}
bam_map = {}
for id, doc in docstore.filter(visible=True, state="ok", name=["OUTPUT_BAM_1", "OUTPUT_BAM_2", "OUTPUT_BAM_3"]):
print doc["name"], doc["tags"]
for t in doc["tags"]:
ts = t.split(":")
if ts[0] == "donor":
if ts[1] not in donor_map:
donor_map[ts[1]] = []
donor_map[ts[1]].append(id)
if ts[0] == "original_bam":
bam_map[ts[1]] = id
if not os.path.exists(args.out):
os.mkdir(args.out)
for key, value in bam_map.items():
t = Target(uuid=value)
path = docstore.get_filename(t)
print "%s\t%s" % (value, path)
os.symlink(path, os.path.join(args.out, "MC3." + key + ".bam"))
def action_cp(args):
rg = nebula.warpdrive.RemoteGalaxy(args.url, args.api_key)
if not args.dir:
docstore = from_url(args.dst)
else:
if not os.path.exists(args.dst):
os.mkdir(args.dst)
for hda in rg.get_history_contents(args.src):
if hda['visible']:
if args.filter is None or re.search(args.filter, hda['name']):
if hda['name'] not in args.exclude:
print hda['name']
meta = rg.get_dataset(hda['id'], 'hda')
if args.dir:
dst_path = os.path.join(args.dst, hda['name'])
rg.download(meta['download_url'], dst_path)
else:
meta['id'] = meta['uuid'] #use the glocal id
hda = Target(uuid=meta['uuid'])
docstore.create(hda)
path = docstore.get_filename(hda)
rg.download(meta['download_url'], path)
docstore.update_from_file(hda)
def run_upload(args):
#syn = synapseclient.Synapse()
#syn.login()
docstore = from_url(args.out_base)
donor_map = {}
bam_map = {}
for id, doc in docstore.filter(
visible=True,
state='ok',
name=['OUTPUT_BAM_1', 'OUTPUT_BAM_2', 'OUTPUT_BAM_3']):
print doc['name'], doc['tags']
for t in doc['tags']:
ts = t.split(":")
if ts[0] == 'donor':
if ts[1] not in donor_map:
donor_map[ts[1]] = []
donor_map[ts[1]].append(id)
if ts[0] == 'original_bam':
bam_map[ts[1]] = id
if not os.path.exists(args.out):
os.mkdir(args.out)
for key, value in bam_map.items():
t = Target(uuid=value)
path = docstore.get_filename(t)
print "%s\t%s" % (value, path)
os.symlink(path, os.path.join(args.out, "MC3." + key + ".bam"))
def run_extract(args):
docstore = from_url(args.out_base)
for id, ent in docstore.filter(
file_ext="vcf",
name=[
"muse.vcf",
"pindel.vcf",
"radia.dna-rna.vcf",
"radia.dna.vcf",
"somatic_sniper.vcf",
"varscan.indel.vcf",
"varscan.snp.vcf",
"mutect.vcf",
],
):
t = Target(uuid=ent["id"])
if docstore.size(t) > 0:
donor = None
for e in ent["tags"]:
tmp = e.split(":")
if tmp[0] == "donor":
donor = tmp[1]
if donor is not None:
donor_dir = os.path.join(args.out_dir, donor)
if not os.path.exists(donor_dir):
os.makedirs(donor_dir)
print "Found", donor, ent["name"]
shutil.copy(docstore.get_filename(t), os.path.join(donor_dir, ent["name"]))
def run_errors(args):
doc = from_url(args.out_base)
for id, entry in doc.filter():
if entry.get('state', '') == 'error':
if args.within is None or 'update_time' not in entry or check_within(entry['update_time'], args.within):
print "Dataset", id, entry.get("job", {}).get("tool_id", ""), entry.get('update_time', ''), entry.get("tags", "")
if args.full:
if 'provenance' in entry:
print "tool:", entry['provenance']['tool_id']
print "-=-=-=-=-=-=-"
print entry['job']['stdout']
print "-------------"
print entry['job']['stderr']
print "-=-=-=-=-=-=-"
def run_errors(args):
doc = from_url(args.out_base)
for id, entry in doc.filter():
if entry.get('state', '') == 'error':
if args.within is None or 'update_time' not in entry or check_within(entry['update_time'], args.within):
print "Dataset", id, entry.get("job", {}).get("tool_id", ""), entry.get('update_time', ''), entry.get("tags", "")
if args.full:
if 'provenance' in entry:
print "tool:", entry['provenance']['tool_id']
print "-=-=-=-=-=-=-"
print entry['job']['stdout']
print "-------------"
print entry['job']['stderr']
print "-=-=-=-=-=-=-"
def run_errors(args):
doc = from_url(args.out_base)
for id, entry in doc.filter():
if entry.get("state", "") == "error":
if args.within is None or "update_time" not in entry or check_within(entry["update_time"], args.within):
print "Dataset", id, entry.get("job", {}).get("tool_id", ""), entry.get("update_time", ""), entry.get(
"tags", ""
)
if args.full:
if "provenance" in entry:
print "tool:", entry["provenance"]["tool_id"]
print "-=-=-=-=-=-=-"
print entry["job"]["stdout"]
print "-------------"
print entry["job"]["stderr"]
print "-=-=-=-=-=-=-"
def run_extract(args):
docstore = from_url(args.out_base)
for id, ent in docstore.filter(file_ext="vcf", name=[
"muse.vcf", "pindel.vcf", "radia.dna-rna.vcf", "radia.dna.vcf", "somatic_sniper.vcf",
"varscan.indel.vcf", "varscan.snp.vcf", "mutect.vcf"
]):
t = Target(uuid=ent['id'])
if docstore.size(t) > 0:
donor = None
for e in ent['tags']:
tmp = e.split(":")
if tmp[0] == 'donor':
donor = tmp[1]
if donor is not None:
donor_dir = os.path.join(args.out_dir, donor)
if not os.path.exists(donor_dir):
os.makedirs(donor_dir)
print "Found", donor, ent['name']
shutil.copy( docstore.get_filename(t), os.path.join(donor_dir, ent['name']) )
def run_upload(args):
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
donor_map = {}
bam_map = {}
for id, doc in docstore.filter(visible=True, state='ok', name=['OUTPUT_BAM_1', 'OUTPUT_BAM_2', 'OUTPUT_BAM_3']):
print doc['name'], doc['tags']
for t in doc['tags']:
ts = t.split(":")
if ts[0] == 'donor':
if ts[1] not in donor_map:
donor_map[ts[1]] = []
donor_map[ts[1]].append(id)
if ts[0] == 'original_bam':
bam_map[ts[1]] = id
print donor_map
print bam_map
def run_audit(args):
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
donor_map = {}
for id, ent in docstore.filter( state="ok" ):
if ent['visible']:
if docstore.size(Target(id)) > 0:
donor = None
for i in ent['tags']:
t = i.split(":")
if t[0] == "donor":
donor = t[1]
if donor not in donor_map:
donor_map[donor] = {}
donor_map[donor][ent['name']] = id
for ent in synqueue.listAssignments(syn, list_all=True, **config):
if ent['meta']['Submitter_donor_ID'] in donor_map:
print ent['meta']['Submitter_donor_ID'], len(donor_map[ent['meta']['Submitter_donor_ID']])
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"db_snp": "dbsnp_132_b37.leftAligned.vcf",
"centromere": "centromere_hg19.bed",
"cosmic": "b37_cosmic_v54_120711.vcf",
}
ref_genomes = [
"Homo_sapiens_assembly19.fasta",
"GRCh37-lite.fa",
"GRCh37-lite-+-HPV_Redux-build.fa",
"GRCh37-lite_WUGSC_variant_1.fa.gz",
"GRCh37-lite_WUGSC_variant_2.fa.gz",
"hg19_M_rCRS.fa.gz",
]
if args.ref_download:
syn_sync(syn, REFDATA_PROJECT, docstore, data_mapping.values() + ref_genomes)
dm = {}
for k, v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = {"uuid": hit}
mc3_dna_workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_CGHub_DNA.ga")
mc3_dnarna_workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_CGHub_DNA_RNA.ga")
rna_hit = None
for a in docstore.filter(name="hg19_M_rCRS.fa"):
rna_hit = a[0]
tasks = TaskGroup()
assembly_hits = {}
with open(args.joblist) as handle:
reader = csv.DictReader(handle, delimiter="\t")
for row in reader:
if row["normal_assembly"] != row["tumor_assembly"]:
print "Row Mispatch", row["normal_assembly"], row["tumor_assembly"]
# raise Exception("Mismatch reference")
ref_name = row["normal_assembly"]
if ref_name in ref_rename:
ref_name = ref_rename[ref_name]
if ref_name in assembly_hits:
hit = assembly_hits[ref_name]
else:
hit = None
for a in docstore.filter(name=ref_name + ".fasta"):
hit = a[0]
for a in docstore.filter(name=ref_name + ".fa"):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (ref_name))
assembly_hits[ref_name] = hit
workflow_dm = dict(dm)
workflow_dm["reference_genome"] = {"uuid": hit}
params = {
"tumor_bam": {
"uuid": row["tumor_analysis_id"],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key",
},
"normal_bam": {
"uuid": row["normal_analysis_id"],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key",
},
"reheader_config": {
"platform": "Illumina",
"center": "OHSU",
"reference_genome": ref_name,
"participant_uuid": row["participant_id"],
"disease_code": row["disease"],
"filedate": datetime.datetime.now().strftime("%Y%m%d"),
"normal_analysis_uuid": row["normal_analysis_id"],
"normal_bam_name": row["normal_filename"],
"normal_aliquot_uuid": row["normal_aliquot_id"],
"normal_aliquot_barcode": row["normal_barcode"],
"tumor_analysis_uuid": row["tumor_analysis_id"],
"tumor_bam_name": row["tumor_filename"],
"tumor_aliquot_uuid": row["tumor_aliquot_id"],
"tumor_aliquot_barcode": row["tumor_barcode"],
},
}
if row["rna_analysis_id"] != "NA":
params["rna_tumor_bam"] = {
"uuid": row["rna_analysis_id"],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key",
}
workflow_dm["rna_reference_genome"] = {"uuid": rna_hit}
task = GalaxyWorkflowTask(
"workflow_%s" % (row["job_id"]),
mc3_dnarna_workflow,
inputs=workflow_dm,
parameters=params,
tags=["donor:%s" % (row["participant_id"])],
)
else:
task = GalaxyWorkflowTask(
"workflow_%s" % (row["job_id"]),
mc3_dna_workflow,
inputs=workflow_dm,
parameters=params,
tags=["donor:%s" % (row["participant_id"])],
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id), "w") as handle:
handle.write(json.dumps(data.to_dict()))
service = GalaxyService(
docstore=docstore,
galaxy=args.galaxy,
sudo=args.sudo,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[
["gatk_bqsr", 12],
["gatk_indel", 24],
["MuSE", 8],
["pindel", 8],
["mutect", 8],
["delly", 4],
["gatk_bqsr", 12],
["gatk_indel", 12],
["bwa_mem", 12],
["radia", 8],
["radia_filter", 8],
],
)
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_uploadprep(args):
if not os.path.exists(args.workdir):
os.mkdir(args.workdir)
doc = from_url(args.out_base)
file_map = {
'broad' : {},
'muse' : {}
}
syn = synapseclient.Synapse()
syn.login()
wl_map = {}
job_map = {}
for ent in synqueue.listAssignments(syn, list_all=True, **config):
wl_map[ent['id']] = ent['meta']
for id, entry in doc.filter():
donor = None
if 'tags' in entry:
for s in entry['tags']:
tmp = s.split(":")
if tmp[0] == 'donor':
donor = tmp[1]
if donor is not None:
if donor not in job_map:
job_map[donor] = {}
if 'job' in entry and 'job_metrics' in entry['job']:
print entry['name']
for met in entry['job']['job_metrics']:
if met['name'] == 'runtime_seconds':
job_map[donor][entry['name']] = {"tool_id" : entry['job']['tool_id'], "runtime_seconds" : met['raw_value']}
if entry.get('visible', False) and entry.get('extension', None) in ["vcf", "vcf_bgzip"]:
pipeline = None
method = None
call_type = None
variant_type = None
if entry['name'].split('.')[0] in ['MUSE_1']:
pipeline = "muse"
method = entry['name'].replace(".", "-")
variant_type = 'somatic'
call_type = 'snv_mnv'
elif entry['name'].split(".")[0] in ['broad-dRanger', 'broad-dRanger_snowman', 'broad-snowman', 'broad-mutect' ]:
pipeline = "broad"
method = entry['name'].split(".")[0]
if 'somatic' in entry['name']:
variant_type = 'somatic'
elif 'germline' in entry['name']:
variant_type = 'germline'
else:
raise Exception("Unknown variant type")
if 'snv_mnv.vcf' in entry['name']:
call_type = 'snv_mnv'
elif 'sv.vcf' in entry['name']:
call_type = 'sv'
elif 'indel.vcf' in entry['name']:
call_type = 'indel'
else:
raise Exception("Unknown call type: %s" % (entry['name']))
else:
raise Exception("Unknown pipeline %s" % (entry['name']))
datestr = datetime.datetime.now().strftime("%Y%m%d")
name = "%s.%s.%s.%s.%s" % (donor, method, datestr, variant_type, call_type )
name = re.sub(r'.vcf$', '', name)
if entry['extension'] == 'vcf':
file_name = name + ".vcf"
elif entry['extension'] == 'vcf_bgzip':
file_name = name + ".vcf.gz"
target = Target(uuid=entry['uuid'])
if doc.size(target) > 0:
src_file = doc.get_filename(target)
dst_file = os.path.join(args.workdir, file_name)
shutil.copy(src_file, dst_file)
if entry['extension'] == 'vcf':
subprocess.check_call( "bgzip -c %s > %s.gz" % (dst_file, dst_file), shell=True )
dst_file = dst_file + ".gz"
subprocess.check_call("tabix -p vcf %s" % (dst_file), shell=True)
shutil.move("%s.tbi" % (dst_file), "%s.idx" % (dst_file))
subprocess.check_call("md5sum %s | awk '{print$1}' > %s.md5" % (dst_file, dst_file), shell=True)
subprocess.check_call("md5sum %s.idx | awk '{print$1}' > %s.idx.md5" % (dst_file, dst_file), shell=True)
if donor not in file_map[pipeline]:
file_map[pipeline][donor] = []
input_file = os.path.basename(dst_file)
file_map[pipeline][donor].append(input_file)
for pipeline, donors in file_map.items():
for donor, files in donors.items():
if donor in wl_map:
"""
with open( os.path.join(args.workdir, "%s.%s.pipeline.json" %(pipeline, donor)), "w" ) as handle:
handle.write(json.dumps( {"pipeline_src" : args.pipeline_src, "pipeline_version" : args.pipeline_version} ))
"""
timing_json = os.path.join(args.workdir, "%s.%s.timing.json" %(pipeline, donor))
with open( timing_json, "w" ) as handle:
handle.write(json.dumps( job_map[donor] ) )
with open( os.path.join(args.workdir, "%s.%s.sh" %(pipeline, donor)), "w" ) as handle:
input_file = os.path.basename(dst_file)
urls = [
"%scghub/metadata/analysisFull/%s" % (wl_map[donor]['Normal_WGS_alignment_GNOS_repos'], wl_map[donor]['Normal_WGS_alignment_GNOS_analysis_ID']),
"%scghub/metadata/analysisFull/%s" % (wl_map[donor]['Tumour_WGS_alignment_GNOS_repos'], wl_map[donor]['Tumour_WGS_alignment_GNOS_analysis_IDs'])
]
cmd_str = "perl /opt/vcf-uploader/gnos_upload_vcf.pl"
cmd_str += " --metadata-urls %s" % (",".join(urls))
cmd_str += " --vcfs %s " % (",".join(files))
cmd_str += " --vcf-md5sum-files %s " % ((",".join( ("%s.md5" % i for i in files) )))
cmd_str += " --vcf-idxs %s" % ((",".join( ("%s.idx" % i for i in files) )))
cmd_str += " --vcf-idx-md5sum-files %s" % ((",".join( ("%s.idx.md5" % i for i in files) )))
cmd_str += " --outdir %s.%s.dir" % (pipeline, donor)
cmd_str += " --key %s " % (args.keyfile)
cmd_str += " --upload-url %s" % (args.upload_url)
cmd_str += " --study-refname-override tcga_pancancer_vcf_test"
cmd_str += " --workflow-src-url '%s'" % args.pipeline_src
cmd_str += " --timing-metrics-json %s" % (timing_json)
handle.write("#!/bin/bash\n%s\n" % (cmd_str) )
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("--out-base", default="pcawg")
parser.add_argument("--ref-download", action="store_true", default=False)
parser.add_argument("--create-service", action="store_true", default=False)
parser.add_argument("--pilot", action="store_true", default=False)
parser.add_argument("--scratch", default=None)
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' % (REFDATA_PROJECT)):
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" % (ent.path, path), shell=True)
name = name.replace(".gz", "")
def run_gen(args):
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"reference_genome" : "genome.fa",
"dbsnp" : "dbsnp_132_b37.leftAligned.vcf",
"cosmic" : "b37_cosmic_v54_120711.vcf",
"gold_indels" : "Mills_and_1000G_gold_standard.indels.hg19.sites.fixed.vcf",
"phase_one_indels" : "1000G_phase1.indels.hg19.sites.fixed.vcf",
"centromere" : "centromere_hg19.bed"
}
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' % (REFDATA_PROJECT)):
print "found", a['entity.name']
if a['entity.name'] in data_mapping.values() or a['entity.name'].replace(".gz", "") in data_mapping.values():
print "loading"
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" % (ent.path, path), shell=True)
name = name.replace(".gz", "")
else:
print "Unknown DataPrep"
else:
shutil.copy(ent.path, path)
docstore.update_from_file(t)
meta = {}
meta['name'] = name
meta['uuid'] = id
if 'dataPrep' in meta:
del meta['dataPrep']
docstore.put(id, meta)
dm = {}
for k,v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = { "uuid" : hit }
workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-PCAWG_CGHUB.ga")
tasks = TaskGroup()
for ent in synqueue.listAssignments(syn, **config):
#print "'%s'" % (ent['state']), ent['state'] == 'nan', type(ent['state']), type('nan')
if not isinstance(ent['state'], basestring) and isnan(ent['state']):
gnos_endpoint = urlparse(ent['meta']['Normal_WGS_alignment_GNOS_repos']).netloc
task = GalaxyWorkflowTask("workflow_%s" % (ent['id']),
workflow,
inputs=dm,
parameters={
'normal_bam_download' : {
"uuid" : ent['meta']['Normal_WGS_alignment_GNOS_analysis_ID'],
"gnos_endpoint" : gnos_endpoint,
"cred_file" : key_map[gnos_endpoint]
},
'tumor_bam_download' : {
"uuid" : ent['meta']['Tumour_WGS_alignment_GNOS_analysis_IDs'],
"gnos_endpoint" : gnos_endpoint,
"cred_file" : key_map[gnos_endpoint]
},
'broad_variant_pipeline' : {
"broad_ref_dir" : "/tool_data/files/refdata",
"sample_id" : ent['meta']['Submitter_donor_ID']
}
},
tags=[ "donor:%s" % (ent['meta']['Submitter_donor_ID']) ]
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id), "w") as handle:
handle.write(json.dumps(data.to_dict()))
state_file = "%s.tasks/%s.state" % (args.out_base, data.task_id)
if os.path.exists( state_file ):
os.unlink( state_file )
print "Tasks Created: %s" % (len(tasks))
if args.create_service:
service = GalaxyService(
docstore=docstore,
galaxy="bgruening/galaxy-stable",
sudo=args.sudo,
tool_data=os.path.abspath("tool_data"),
tool_dir=os.path.abspath("tools"),
work_dir=args.work_dir,
smp=[
["MuSE", 8],
["pindel", 8],
["muTect", 8],
["delly", 4],
["gatk_bqsr", 12],
["gatk_indel", 24],
["bwa_mem", 12],
["broad_variant_pipline", 24]
]
)
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"db_snp" : "dbsnp_132_b37.leftAligned.vcf",
"centromere" : "centromere_hg19.bed",
"cosmic" : "b37_cosmic_v54_120711.vcf"
}
ref_genomes = [
"Homo_sapiens_assembly19.fasta",
"GRCh37-lite.fa",
"GRCh37-lite-+-HPV_Redux-build.fa",
"GRCh37-lite_WUGSC_variant_1.fa.gz",
"GRCh37-lite_WUGSC_variant_2.fa.gz",
"hg19_M_rCRS.fa.gz"
]
if args.ref_download:
syn_sync(syn, REFDATA_PROJECT, docstore, data_mapping.values() + ref_genomes)
dm = {}
for k,v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = { "uuid" : hit }
mc3_dna_workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_CGHub_DNA.ga")
mc3_dnarna_workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_CGHub_DNA_RNA.ga")
rna_hit = None
for a in docstore.filter(name="hg19_M_rCRS.fa"):
rna_hit = a[0]
tasks = TaskGroup()
assembly_hits = {}
with open(args.joblist) as handle:
reader = csv.DictReader(handle, delimiter="\t")
for row in reader:
if row['normal_assembly'] != row['tumor_assembly']:
print "Row Mispatch", row['normal_assembly'], row['tumor_assembly']
#raise Exception("Mismatch reference")
ref_name = row['normal_assembly']
if ref_name in ref_rename:
ref_name = ref_rename[ref_name]
if ref_name in assembly_hits:
hit = assembly_hits[ref_name]
else:
hit = None
for a in docstore.filter(name=ref_name + ".fasta"):
hit = a[0]
for a in docstore.filter(name=ref_name + ".fa"):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (ref_name))
assembly_hits[ref_name] = hit
workflow_dm = dict(dm)
workflow_dm['reference_genome'] = { "uuid" : hit }
params = {
'tumor_bam' : {
"uuid" : row['tumor_analysis_id'],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
},
'normal_bam' : {
"uuid" : row['normal_analysis_id'],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
},
"reheader_config" : {
"platform" : "Illumina",
"center" : "OHSU",
"reference_genome" : ref_name,
"participant_uuid" : row['participant_id'],
"disease_code" : row['disease'],
"filedate" : datetime.datetime.now().strftime("%Y%m%d"),
"normal_analysis_uuid" : row['normal_analysis_id'],
"normal_bam_name" : row['normal_filename'],
"normal_aliquot_uuid" : row['normal_aliquot_id'],
"normal_aliquot_barcode": row['normal_barcode'],
"tumor_analysis_uuid" : row['tumor_analysis_id'],
"tumor_bam_name" : row['tumor_filename'],
"tumor_aliquot_uuid" : row['tumor_aliquot_id'],
"tumor_aliquot_barcode" : row['tumor_barcode'],
}
}
if row['rna_analysis_id'] != "NA":
params['rna_tumor_bam'] = {
"uuid" : row['rna_analysis_id'],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
}
workflow_dm['rna_reference_genome'] = { "uuid" : rna_hit }
task = GalaxyWorkflowTask("workflow_%s" % (row['job_id']),
mc3_dnarna_workflow,
inputs=workflow_dm,
parameters=params,
tags=[ "donor:%s" % (row['participant_id']) ],
)
else:
task = GalaxyWorkflowTask("workflow_%s" % (row['job_id']),
mc3_dna_workflow,
inputs=workflow_dm,
parameters=params,
tags=[ "donor:%s" % (row['participant_id']) ],
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id), "w") as handle:
handle.write(json.dumps(data.to_dict()))
service = GalaxyService(
docstore=docstore,
galaxy=args.galaxy,
sudo=args.sudo,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[
["gatk_bqsr", 12],
["gatk_indel", 24],
["MuSE", 8],
["pindel", 8],
["mutect", 8],
["delly", 4],
["gatk_bqsr", 12],
["gatk_indel", 12],
["bwa_mem", 12],
["radia", 8],
['radia_filter', 8]
]
)
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
if args.alt_table is not None:
config['table_id'] = args.alt_table
docstore = from_url(args.out_base)
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' %
(REFDATA_PROJECT)):
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" %
(ent.path, path),
shell=True)
name = name.replace(".gz", "")
else:
print "Unknown DataPrep"
else:
shutil.copy(ent.path, path)
docstore.update_from_file(t)
meta = {}
meta['name'] = name
meta['uuid'] = id
if 'dataPrep' in meta:
del meta['dataPrep']
docstore.put(id, meta)
data_mapping = {
"dbsnp": "dbsnp_132_b37.leftAligned.vcf",
"cosmic": "b37_cosmic_v54_120711.vcf",
"gold_indels":
"Mills_and_1000G_gold_standard.indels.hg19.sites.fixed.vcf",
"phase_one_indels": "1000G_phase1.indels.hg19.sites.fixed.vcf"
}
dm = {}
for k, v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = {"uuid": hit}
workflow_2 = GalaxyWorkflow(
ga_file="workflows/Galaxy-Workflow-GATK_CGHub_2.ga")
workflow_3 = GalaxyWorkflow(
ga_file="workflows/Galaxy-Workflow-GATK_CGHub_3.ga")
ref_rename = {"HG19_Broad_variant": "Homo_sapiens_assembly19"}
tasks = TaskGroup()
for ent in synqueue.listAssignments(syn, **config):
bam_set = list(
a[1] for a in ent['meta'].items()
if a[0].startswith("id_") and isinstance(a[1], basestring))
ref_set = set(a[1] for a in ent['meta'].items()
if a[0].startswith("ref_assembly_")
and isinstance(a[1], basestring))
assert (len(ref_set) == 1)
ref_name = ref_set.pop()
if ref_name in ref_rename:
ref_name = ref_rename[ref_name]
hit = None
for a in docstore.filter(name=ref_name + ".fasta"):
hit = a[0]
for a in docstore.filter(name=ref_name + ".fa"):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (ref_name))
workflow_dm = dict(dm)
workflow_dm['reference_genome'] = {"uuid": hit}
if len(bam_set) == 2:
task = GalaxyWorkflowTask(
"workflow_%s" % (ent['id']),
workflow_2,
inputs=workflow_dm,
parameters={
'INPUT_BAM_1': {
"uuid": bam_set[0],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key"
},
'INPUT_BAM_2': {
"uuid": bam_set[1],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key"
}
},
tags=["donor:%s" % (ent['meta']['participant_id'])],
tool_tags={
"BQSR_1": {
"output_bam": ["original_bam:%s" % (bam_set[0])]
},
"BQSR_2": {
"output_bam": ["original_bam:%s" % (bam_set[1])]
}
})
tasks.append(task)
elif len(bam_set) == 3:
task = GalaxyWorkflowTask(
"workflow_%s" % (ent['id']),
workflow_3,
inputs=workflow_dm,
parameters={
'INPUT_BAM_1': {
"uuid": bam_set[0],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key"
},
'INPUT_BAM_2': {
"uuid": bam_set[1],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key"
},
'INPUT_BAM_3': {
"uuid": bam_set[2],
"gnos_endpoint": "cghub.ucsc.edu",
"cred_file": "/tool_data/files/cghub.key"
}
},
tags=["donor:%s" % (ent['meta']['participant_id'])],
tool_tags={
"BQSR_1": {
"output_bam": ["original_bam:%s" % (bam_set[0])]
},
"BQSR_2": {
"output_bam": ["original_bam:%s" % (bam_set[1])]
},
"BQSR_3": {
"output_bam": ["original_bam:%s" % (bam_set[2])]
}
})
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id),
"w") as handle:
handle.write(json.dumps(data.to_dict()))
if args.create_service:
service = GalaxyService(docstore=docstore,
galaxy="bgruening/galaxy-stable",
sudo=True,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[["gatk_bqsr", 12], ["gatk_indel", 24]])
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
REFDATA_PROJECT="syn3241088"
if __name__ == "__main__":
parser = argparse.ArgumentParser()
parser.add_argument("docstore", help="DocStore")
parser.add_argument("--ref-download", action="store_true", default=False)
parser.add_argument("--create-service", action="store_true", default=False)
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.docstore)
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' % (REFDATA_PROJECT)):
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" % (ent.path, path), shell=True)
name = name.replace(".gz", "")
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
if args.alt_table is not None:
config['table_id'] = args.alt_table
docstore = from_url(args.out_base)
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' % (REFDATA_PROJECT)):
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" % (ent.path, path), shell=True)
name = name.replace(".gz", "")
else:
print "Unknown DataPrep"
else:
shutil.copy(ent.path, path)
docstore.update_from_file(t)
meta = {}
meta['name'] = name
meta['uuid'] = id
if 'dataPrep' in meta:
del meta['dataPrep']
docstore.put(id, meta)
data_mapping = {
"dbsnp" : "dbsnp_132_b37.leftAligned.vcf",
"cosmic" : "b37_cosmic_v54_120711.vcf",
"gold_indels" : "Mills_and_1000G_gold_standard.indels.hg19.sites.fixed.vcf",
"phase_one_indels" : "1000G_phase1.indels.hg19.sites.fixed.vcf"
}
dm = {}
for k,v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = { "uuid" : hit }
workflow_2 = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-GATK_CGHub_2.ga")
workflow_3 = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-GATK_CGHub_3.ga")
ref_rename = {
"HG19_Broad_variant" : "Homo_sapiens_assembly19"
}
tasks = TaskGroup()
for ent in synqueue.listAssignments(syn, **config):
bam_set = list( a[1] for a in ent['meta'].items() if a[0].startswith("id_") and isinstance(a[1], basestring) )
ref_set = set( a[1] for a in ent['meta'].items() if a[0].startswith("ref_assembly_") and isinstance(a[1], basestring) )
assert(len(ref_set) == 1)
ref_name = ref_set.pop()
if ref_name in ref_rename:
ref_name = ref_rename[ref_name]
hit = None
for a in docstore.filter(name=ref_name + ".fasta"):
hit = a[0]
for a in docstore.filter(name=ref_name + ".fa"):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (ref_name))
workflow_dm = dict(dm)
workflow_dm['reference_genome'] = { "uuid" : hit }
if len(bam_set) == 2:
task = GalaxyWorkflowTask("workflow_%s" % (ent['id']),
workflow_2,
inputs=workflow_dm,
parameters={
'INPUT_BAM_1' : {
"uuid" : bam_set[0],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
},
'INPUT_BAM_2' : {
"uuid" : bam_set[1],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
}
},
tags=[ "donor:%s" % (ent['meta']['participant_id']) ],
tool_tags = {
"BQSR_1" : {
"output_bam" : [ "original_bam:%s" % (bam_set[0]) ]
},
"BQSR_2" : {
"output_bam" : [ "original_bam:%s" % (bam_set[1]) ]
}
}
)
tasks.append(task)
elif len(bam_set) == 3:
task = GalaxyWorkflowTask("workflow_%s" % (ent['id']),
workflow_3,
inputs=workflow_dm,
parameters={
'INPUT_BAM_1' : {
"uuid" : bam_set[0],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
},
'INPUT_BAM_2' : {
"uuid" : bam_set[1],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
},
'INPUT_BAM_3' : {
"uuid" : bam_set[2],
"gnos_endpoint" : "cghub.ucsc.edu",
"cred_file" : "/tool_data/files/cghub.key"
}
},
tags=[ "donor:%s" % (ent['meta']['participant_id']) ],
tool_tags = {
"BQSR_1" : {
"output_bam" : [ "original_bam:%s" % (bam_set[0]) ]
},
"BQSR_2" : {
"output_bam" : [ "original_bam:%s" % (bam_set[1]) ]
},
"BQSR_3" : {
"output_bam" : [ "original_bam:%s" % (bam_set[2]) ]
}
}
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id), "w") as handle:
handle.write(json.dumps(data.to_dict()))
if args.create_service:
service = GalaxyService(
docstore=docstore,
galaxy="bgruening/galaxy-stable",
sudo=True,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[
["gatk_bqsr", 12],
["gatk_indel", 24]
]
)
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_uploadprep(args):
if not os.path.exists(args.workdir):
os.mkdir(args.workdir)
doc = from_url(args.out_base)
file_map = {
'broad' : {},
'muse' : {},
'broad_tar' : {}
}
syn = synapseclient.Synapse()
syn.login()
wl_map = {}
job_map = {}
for ent in synqueue.listAssignments(syn, list_all=True, **config):
wl_map[ent['id']] = ent['meta']
uuid_map = {}
uuid_map['broad'] = synqueue.getValues(syn, "Broad_VCF_UUID", orSet=lambda x: str(uuid.uuid4()), **config)
uuid_map['muse'] = synqueue.getValues(syn, "Muse_VCF_UUID", orSet=lambda x: str(uuid.uuid4()), **config)
uuid_map['broad_tar'] = synqueue.getValues(syn, "Broad_TAR_UUID", orSet=lambda x: str(uuid.uuid4()), **config)
#scan through all of the docs
for id, entry in doc.filter():
donor = None
#look for docs with donor tags
if 'tags' in entry and 'state' in entry and entry['state'] == 'ok':
for s in entry['tags']:
tmp = s.split(":")
if tmp[0] == 'donor':
donor = tmp[1]
if donor is not None and donor in wl_map:
if donor not in job_map:
job_map[donor] = {}
#scan out the job metrics for this job
if 'job' in entry and 'job_metrics' in entry['job']:
job_id = entry['job']['id']
tool_id = entry['job']['tool_id']
job_info = { tool_id : {} }
for met in entry['job']['job_metrics']:
job_info[tool_id][met['name']] = met['raw_value']
job_map[donor][job_id] = job_info
donor_tumor = wl_map[donor]['Tumour_WGS_alignment_GNOS_analysis_IDs']
#look for the vcf output files
if entry.get('visible', False) and entry.get('extension', None) in ["vcf", "vcf_bgzip"]:
pipeline = None
method = None
call_type = None
variant_type = None
#fill out the info depending on which caller created the file
if entry['name'].split('.')[0] in ['MUSE_1']:
pipeline = "muse"
method = entry['name'].replace(".", "-")
variant_type = 'somatic'
call_type = 'snv_mnv'
elif entry['name'].split(".")[0] in ['broad-dRanger', 'broad-dRanger_snowman', 'broad-snowman', 'broad-mutect' ]:
pipeline = "broad"
method = entry['name'].split(".")[0]
if 'somatic' in entry['name']:
variant_type = 'somatic'
elif 'germline' in entry['name']:
variant_type = 'germline'
else:
raise Exception("Unknown variant type")
if 'snv_mnv.vcf' in entry['name']:
call_type = 'snv_mnv'
elif 'sv.vcf' in entry['name']:
call_type = 'sv'
elif 'indel.vcf' in entry['name']:
call_type = 'indel'
else:
raise Exception("Unknown call type: %s" % (entry['name']))
else:
raise Exception("Unknown pipeline %s" % (entry['name']))
datestr = datetime.datetime.now().strftime("%Y%m%d")
name = "%s.%s.%s.%s.%s" % (donor_tumor, method, datestr, variant_type, call_type )
name = re.sub(r'.vcf$', '', name)
if entry['extension'] == 'vcf':
file_name = name + ".vcf"
elif entry['extension'] == 'vcf_bgzip':
file_name = name + ".vcf.gz"
target = Target(uuid=entry['uuid'])
if doc.size(target) > 0:
src_file = doc.get_filename(target)
dst_file = os.path.join(args.workdir, file_name)
shutil.copy(src_file, dst_file)
#if the files wasn't compressed already, go ahead and do that
if entry['extension'] == 'vcf':
subprocess.check_call( "bgzip -c %s > %s.gz" % (dst_file, dst_file), shell=True )
dst_file = dst_file + ".gz"
#add file to output map
if donor not in file_map[pipeline]:
file_map[pipeline][donor] = []
input_file = os.path.basename(dst_file)
file_map[pipeline][donor].append(input_file)
else:
if entry['name'] == "broad.tar.gz":
target = Target(uuid=entry['uuid'])
src_file = doc.get_filename(target)
file_map['broad_tar'][donor] = [ src_file ]
timing_map = {}
for donor in job_map:
timing_map[donor] = {}
for job_id in job_map[donor]:
for tool_id in job_map[donor][job_id]:
if tool_id not in timing_map[donor]:
timing_map[donor][tool_id] = []
timing_map[donor][tool_id].append( job_map[donor][job_id][tool_id] )
result_counts = {}
for pipeline, donors in file_map.items():
for donor in donors:
result_counts[donor] = result_counts.get(donor, 0) + 1
#go through every pipeline
for pipeline, donors in file_map.items():
#for that pipeline go through every donor
for donor, files in donors.items():
#we're only outputing data for donors on the work list
if donor in wl_map and result_counts[donor] == 3:
#output the timing json
timing_json = os.path.abspath(os.path.join(args.workdir, "%s.%s.timing.json" %(pipeline, donor)))
with open( timing_json, "w" ) as handle:
handle.write(json.dumps( timing_map[donor] ) )
#output the uploader script
with open( os.path.join(args.workdir, "%s.%s.sh" %(pipeline, donor)), "w" ) as handle:
input_file = os.path.basename(dst_file)
urls = [
"%scghub/metadata/analysisFull/%s" % (wl_map[donor]['Normal_WGS_alignment_GNOS_repos'], wl_map[donor]['Normal_WGS_alignment_GNOS_analysis_ID']),
"%scghub/metadata/analysisFull/%s" % (wl_map[donor]['Tumour_WGS_alignment_GNOS_repos'], wl_map[donor]['Tumour_WGS_alignment_GNOS_analysis_IDs'])
]
donor_tumor = wl_map[donor]['Tumour_WGS_alignment_GNOS_analysis_IDs']
if pipeline in ['broad', 'muse']:
prep_cmd_str = ""
for vcf in files:
prep_cmd_str += "tabix -p vcf %s\n" % (vcf)
prep_cmd_str += "mv %s.tbi %s.idx\n" % (vcf,vcf)
prep_cmd_str += "md5sum %s | awk '{print$1}' > %s.md5\n" % (vcf, vcf)
prep_cmd_str += "md5sum %s.idx | awk '{print$1}' > %s.idx.md5\n\n" % (vcf, vcf)
related_uuids = []
for p in uuid_map:
if p != pipeline:
related_uuids.append(uuid_map[p][donor])
submit_cmd_str = "perl -I /opt/gt-download-upload-wrapper/gt-download-upload-wrapper-2.0.11/lib"
submit_cmd_str += " /opt/vcf-uploader/vcf-uploader-2.0.5/gnos_upload_vcf.pl"
submit_cmd_str += " --metadata-urls %s" % (",".join(urls))
submit_cmd_str += " --vcfs %s " % (",".join(files))
submit_cmd_str += " --vcf-md5sum-files %s " % ((",".join( ("%s.md5" % i for i in files) )))
submit_cmd_str += " --vcf-idxs %s" % ((",".join( ("%s.idx" % i for i in files) )))
submit_cmd_str += " --vcf-idx-md5sum-files %s" % ((",".join( ("%s.idx.md5" % i for i in files) )))
submit_cmd_str += " --outdir %s.%s.dir" % (pipeline, donor_tumor)
submit_cmd_str += " --key %s " % (args.keyfile)
submit_cmd_str += " --upload-url %s" % (args.upload_url)
submit_cmd_str += " --study-refname-override %s" % (args.study)
submit_cmd_str += " --workflow-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-src-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-name '%s'" % args.pipeline_name
submit_cmd_str += " --workflow-version '%s'" % args.pipeline_version
submit_cmd_str += " --vm-instance-type '%s'" % args.vm_instance_type
submit_cmd_str += " --vm-instance-cores %s" % args.vm_instance_cores
submit_cmd_str += " --vm-instance-mem-gb %s" % args.vm_instance_mem_gb
submit_cmd_str += " --vm-location-code %s" % args.vm_location_code
submit_cmd_str += " --timing-metrics-json %s" % (timing_json)
submit_cmd_str += " --workflow-file-subset %s" % (pipeline)
submit_cmd_str += " --related-file-subset-uuids %s" % (",".join(related_uuids))
submit_cmd_str += " --uuid %s" % (uuid_map[pipeline][donor])
#submit_cmd_str += " --skip-upload"
if pipeline in ['broad_tar']:
prep_cmd_str = ""
new_files = []
for tar in files:
basename = donor_tumor + ".broad.intermediate"
prep_cmd_str = "%s/remap_broad_tar.py %s %s %s --rename %s %s" % (
os.path.dirname(os.path.abspath(__file__)),
tar,
"./",
basename,
donor, donor_tumor
)
new_files.append( basename + ".tar" )
related_uuids = []
for p in uuid_map:
if p != pipeline:
related_uuids.append(uuid_map[p][donor])
submit_cmd_str = "perl -I /opt/gt-download-upload-wrapper/gt-download-upload-wrapper-2.0.11/lib"
submit_cmd_str += " /opt/vcf-uploader/vcf-uploader-2.0.5/gnos_upload_vcf.pl"
submit_cmd_str += " --metadata-urls %s" % (",".join(urls))
submit_cmd_str += " --tarballs %s " % (",".join(new_files))
submit_cmd_str += " --tarball-md5sum-files %s " % ((",".join( ("%s.md5" % i for i in new_files) )))
submit_cmd_str += " --outdir %s.%s.dir" % (pipeline, donor_tumor)
submit_cmd_str += " --key %s " % (args.keyfile)
submit_cmd_str += " --upload-url %s" % (args.upload_url)
submit_cmd_str += " --study-refname-override %s" % (args.study)
submit_cmd_str += " --workflow-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-src-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-name '%s'" % args.pipeline_name
submit_cmd_str += " --workflow-version '%s'" % args.pipeline_version
submit_cmd_str += " --vm-instance-type '%s'" % args.vm_instance_type
submit_cmd_str += " --vm-instance-cores %s" % args.vm_instance_cores
submit_cmd_str += " --vm-instance-mem-gb %s" % args.vm_instance_mem_gb
submit_cmd_str += " --workflow-file-subset %s" % (pipeline)
submit_cmd_str += " --timing-metrics-json %s" % (timing_json)
submit_cmd_str += " --related-file-subset-uuids %s" % (",".join(related_uuids))
submit_cmd_str += " --uuid %s" % (uuid_map[pipeline][donor])
#submit_cmd_str += " --skip-upload"
handle.write(string.Template("""#!/bin/bash
set -ex
${PREP}
${SUBMIT}
echo $$? > $$0.submitted
#pushd ${SUBMIT_DIR}
#gtupload -v -c ${KEY} -u ./manifest.xml
#ECODE=$$?
#popd
#echo $$ECODE > $$0.uploaded
""").substitute(PREP=prep_cmd_str, SUBMIT=submit_cmd_str,
SUBMIT_DIR=os.path.join(os.path.abspath(args.workdir), "vcf", pipeline + "." + donor_tumor + ".dir", uuid_map[pipeline][donor] ),
KEY=args.keyfile
) )
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"db_snp" : "dbsnp_132_b37.leftAligned.vcf",
"centromere" : "centromere_hg19.bed",
"reference_genome" : "Homo_sapiens_assembly19.fasta",
"cosmic" : "b37_cosmic_v54_120711.vcf"
}
if args.ref_download:
syn_sync(syn, REFDATA_PROJECT, docstore, data_mapping.values())
dm = {}
for k,v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = { "uuid" : hit }
if args.sample is not None:
sync_doc_dir(
os.path.join( os.path.dirname(__file__), "..", "testexomes" ), docstore,
filter=lambda x: x['donorId'] in args.sample
)
else:
sync_doc_dir( os.path.join( os.path.dirname(__file__), "..", "testexomes" ), docstore)
tumor_uuids = {}
normal_uuids = {}
for id, ent in docstore.filter(sampleType="tumour"):
tumor_uuids[ent['participant_id']] = id
for id, ent in docstore.filter(sampleType="normal"):
normal_uuids[ent['participant_id']] = id
mc3_workflow = GalaxyWorkflow(ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_Test.ga")
reference_id = None
for a in docstore.filter(name="Homo_sapiens_assembly19.fasta"):
reference_id = a[0]
tasks = TaskGroup()
for donor in tumor_uuids:
if donor in normal_uuids:
print "participant", donor
donor_name = None
for k,v in fake_metadata.items():
if v['participant_id'] == donor:
donor_name = k
workflow_dm = dict(dm)
workflow_dm['tumor_bam'] = { "uuid" : tumor_uuids[donor] }
workflow_dm['normal_bam'] = { "uuid" : normal_uuids[donor] }
task = GalaxyWorkflowTask("workflow_%s" % (donor),
mc3_workflow,
inputs=workflow_dm,
parameters={
"reheader_config" : {
"platform" : "Illumina",
"center" : "OHSU",
"reference_genome" : "Homo_sapiens_assembly19.fasta",
"participant_uuid" : fake_metadata[donor_name]['participant_id'],
"disease_code" : fake_metadata[donor_name]['disease'],
"filedate" : datetime.datetime.now().strftime("%Y%m%d"),
"normal_analysis_uuid" : fake_metadata[donor_name]['normal']['uuid'],
"normal_bam_name" : fake_metadata[donor_name]['normal']['file_name'],
"normal_aliquot_uuid" : fake_metadata[donor_name]['normal']['aliquot_id'],
"normal_aliquot_barcode": fake_metadata[donor_name]['normal']['barcode'],
"tumor_analysis_uuid" : fake_metadata[donor_name]['tumour']['uuid'],
"tumor_bam_name" : fake_metadata[donor_name]['tumour']['file_name'],
"tumor_aliquot_uuid" : fake_metadata[donor_name]['tumour']['aliquot_id'],
"tumor_aliquot_barcode" : fake_metadata[donor_name]['tumour']['barcode'],
}
},
tags=[ "donor:%s" % (donor) ],
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id), "w") as handle:
handle.write(json.dumps(data.to_dict()))
if args.create_service:
service = GalaxyService(
docstore=docstore,
galaxy=args.galaxy,
sudo=args.sudo,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[
["gatk_bqsr", 12],
["gatk_indel", 24],
["MuSE", 8],
["pindel", 8],
["mutect", 8],
["delly", 4],
["gatk_bqsr", 12],
["gatk_indel", 12],
["bwa_mem", 12],
["radia", 8],
['radia_filter', 8]
]
)
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_uploadprep(args):
if not os.path.exists(args.workdir):
os.mkdir(args.workdir)
doc = from_url(args.out_base)
file_map = {'broad': {}, 'muse': {}, 'broad_tar': {}}
syn = synapseclient.Synapse()
syn.login()
wl_map = {}
job_map = {}
for ent in synqueue.listAssignments(syn, list_all=True, **config):
wl_map[ent['id']] = ent['meta']
uuid_map = {}
uuid_map['broad'] = synqueue.getValues(syn,
"Broad_VCF_UUID",
orSet=lambda x: str(uuid.uuid4()),
**config)
uuid_map['muse'] = synqueue.getValues(syn,
"Muse_VCF_UUID",
orSet=lambda x: str(uuid.uuid4()),
**config)
uuid_map['broad_tar'] = synqueue.getValues(
syn, "Broad_TAR_UUID", orSet=lambda x: str(uuid.uuid4()), **config)
#scan through all of the docs
for id, entry in doc.filter():
donor = None
#look for docs with donor tags
if 'tags' in entry and 'state' in entry and entry['state'] == 'ok':
for s in entry['tags']:
tmp = s.split(":")
if tmp[0] == 'donor':
donor = tmp[1]
if donor is not None and donor in wl_map:
if donor not in job_map:
job_map[donor] = {}
#scan out the job metrics for this job
if 'job' in entry and 'job_metrics' in entry['job']:
job_id = entry['job']['id']
tool_id = entry['job']['tool_id']
job_info = {tool_id: {}}
for met in entry['job']['job_metrics']:
job_info[tool_id][met['name']] = met['raw_value']
job_map[donor][job_id] = job_info
donor_tumor = wl_map[donor][
'Tumour_WGS_alignment_GNOS_analysis_IDs']
#look for the vcf output files
if entry.get('visible', False) and entry.get(
'extension', None) in ["vcf", "vcf_bgzip"]:
pipeline = None
method = None
call_type = None
variant_type = None
#fill out the info depending on which caller created the file
if entry['name'].split('.')[0] in ['MUSE_1']:
pipeline = "muse"
method = entry['name'].replace(".", "-")
variant_type = 'somatic'
call_type = 'snv_mnv'
elif entry['name'].split(".")[0] in [
'broad-dRanger', 'broad-dRanger_snowman',
'broad-snowman', 'broad-mutect'
]:
pipeline = "broad"
method = entry['name'].split(".")[0]
if 'somatic' in entry['name']:
variant_type = 'somatic'
elif 'germline' in entry['name']:
variant_type = 'germline'
else:
raise Exception("Unknown variant type")
if 'snv_mnv.vcf' in entry['name']:
call_type = 'snv_mnv'
elif 'sv.vcf' in entry['name']:
call_type = 'sv'
elif 'indel.vcf' in entry['name']:
call_type = 'indel'
else:
raise Exception("Unknown call type: %s" %
(entry['name']))
else:
raise Exception("Unknown pipeline %s" % (entry['name']))
datestr = datetime.datetime.now().strftime("%Y%m%d")
name = "%s.%s.%s.%s.%s" % (donor_tumor, method, datestr,
variant_type, call_type)
name = re.sub(r'.vcf$', '', name)
if entry['extension'] == 'vcf':
file_name = name + ".vcf"
elif entry['extension'] == 'vcf_bgzip':
file_name = name + ".vcf.gz"
target = Target(uuid=entry['uuid'])
if doc.size(target) > 0:
src_file = doc.get_filename(target)
dst_file = os.path.join(args.workdir, file_name)
shutil.copy(src_file, dst_file)
#if the files wasn't compressed already, go ahead and do that
if entry['extension'] == 'vcf':
subprocess.check_call("bgzip -c %s > %s.gz" %
(dst_file, dst_file),
shell=True)
dst_file = dst_file + ".gz"
#add file to output map
if donor not in file_map[pipeline]:
file_map[pipeline][donor] = []
input_file = os.path.basename(dst_file)
file_map[pipeline][donor].append(input_file)
else:
if entry['name'] == "broad.tar.gz":
target = Target(uuid=entry['uuid'])
src_file = doc.get_filename(target)
file_map['broad_tar'][donor] = [src_file]
timing_map = {}
for donor in job_map:
timing_map[donor] = {}
for job_id in job_map[donor]:
for tool_id in job_map[donor][job_id]:
if tool_id not in timing_map[donor]:
timing_map[donor][tool_id] = []
timing_map[donor][tool_id].append(
job_map[donor][job_id][tool_id])
result_counts = {}
for pipeline, donors in file_map.items():
for donor in donors:
result_counts[donor] = result_counts.get(donor, 0) + 1
#go through every pipeline
for pipeline, donors in file_map.items():
#for that pipeline go through every donor
for donor, files in donors.items():
#we're only outputing data for donors on the work list
if donor in wl_map and result_counts[donor] == 3:
#output the timing json
timing_json = os.path.abspath(
os.path.join(args.workdir,
"%s.%s.timing.json" % (pipeline, donor)))
with open(timing_json, "w") as handle:
handle.write(json.dumps(timing_map[donor]))
#output the uploader script
with open(
os.path.join(args.workdir,
"%s.%s.sh" % (pipeline, donor)),
"w") as handle:
input_file = os.path.basename(dst_file)
urls = [
"%scghub/metadata/analysisFull/%s" %
(wl_map[donor]['Normal_WGS_alignment_GNOS_repos'],
wl_map[donor]['Normal_WGS_alignment_GNOS_analysis_ID']
),
"%scghub/metadata/analysisFull/%s" %
(wl_map[donor]['Tumour_WGS_alignment_GNOS_repos'],
wl_map[donor]
['Tumour_WGS_alignment_GNOS_analysis_IDs'])
]
donor_tumor = wl_map[donor][
'Tumour_WGS_alignment_GNOS_analysis_IDs']
if pipeline in ['broad', 'muse']:
prep_cmd_str = ""
for vcf in files:
prep_cmd_str += "tabix -p vcf %s\n" % (vcf)
prep_cmd_str += "mv %s.tbi %s.idx\n" % (vcf, vcf)
prep_cmd_str += "md5sum %s | awk '{print$1}' > %s.md5\n" % (
vcf, vcf)
prep_cmd_str += "md5sum %s.idx | awk '{print$1}' > %s.idx.md5\n\n" % (
vcf, vcf)
related_uuids = []
for p in uuid_map:
if p != pipeline:
related_uuids.append(uuid_map[p][donor])
submit_cmd_str = "perl -I /opt/gt-download-upload-wrapper/gt-download-upload-wrapper-2.0.11/lib"
submit_cmd_str += " /opt/vcf-uploader/vcf-uploader-2.0.5/gnos_upload_vcf.pl"
submit_cmd_str += " --metadata-urls %s" % (
",".join(urls))
submit_cmd_str += " --vcfs %s " % (",".join(files))
submit_cmd_str += " --vcf-md5sum-files %s " % (
(",".join(("%s.md5" % i for i in files))))
submit_cmd_str += " --vcf-idxs %s" % ((",".join(
("%s.idx" % i for i in files))))
submit_cmd_str += " --vcf-idx-md5sum-files %s" % (
(",".join(("%s.idx.md5" % i for i in files))))
submit_cmd_str += " --outdir %s.%s.dir" % (pipeline,
donor_tumor)
submit_cmd_str += " --key %s " % (args.keyfile)
submit_cmd_str += " --upload-url %s" % (
args.upload_url)
submit_cmd_str += " --study-refname-override %s" % (
args.study)
submit_cmd_str += " --workflow-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-src-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-name '%s'" % args.pipeline_name
submit_cmd_str += " --workflow-version '%s'" % args.pipeline_version
submit_cmd_str += " --vm-instance-type '%s'" % args.vm_instance_type
submit_cmd_str += " --vm-instance-cores %s" % args.vm_instance_cores
submit_cmd_str += " --vm-instance-mem-gb %s" % args.vm_instance_mem_gb
submit_cmd_str += " --vm-location-code %s" % args.vm_location_code
submit_cmd_str += " --timing-metrics-json %s" % (
timing_json)
submit_cmd_str += " --workflow-file-subset %s" % (
pipeline)
submit_cmd_str += " --related-file-subset-uuids %s" % (
",".join(related_uuids))
submit_cmd_str += " --uuid %s" % (
uuid_map[pipeline][donor])
#submit_cmd_str += " --skip-upload"
if pipeline in ['broad_tar']:
prep_cmd_str = ""
new_files = []
for tar in files:
basename = donor_tumor + ".broad.intermediate"
prep_cmd_str = "%s/remap_broad_tar.py %s %s %s --rename %s %s" % (
os.path.dirname(os.path.abspath(__file__)),
tar, "./", basename, donor, donor_tumor)
new_files.append(basename + ".tar")
related_uuids = []
for p in uuid_map:
if p != pipeline:
related_uuids.append(uuid_map[p][donor])
submit_cmd_str = "perl -I /opt/gt-download-upload-wrapper/gt-download-upload-wrapper-2.0.11/lib"
submit_cmd_str += " /opt/vcf-uploader/vcf-uploader-2.0.5/gnos_upload_vcf.pl"
submit_cmd_str += " --metadata-urls %s" % (
",".join(urls))
submit_cmd_str += " --tarballs %s " % (
",".join(new_files))
submit_cmd_str += " --tarball-md5sum-files %s " % (
(",".join(("%s.md5" % i for i in new_files))))
submit_cmd_str += " --outdir %s.%s.dir" % (pipeline,
donor_tumor)
submit_cmd_str += " --key %s " % (args.keyfile)
submit_cmd_str += " --upload-url %s" % (
args.upload_url)
submit_cmd_str += " --study-refname-override %s" % (
args.study)
submit_cmd_str += " --workflow-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-src-url '%s'" % args.pipeline_src
submit_cmd_str += " --workflow-name '%s'" % args.pipeline_name
submit_cmd_str += " --workflow-version '%s'" % args.pipeline_version
submit_cmd_str += " --vm-instance-type '%s'" % args.vm_instance_type
submit_cmd_str += " --vm-instance-cores %s" % args.vm_instance_cores
submit_cmd_str += " --vm-instance-mem-gb %s" % args.vm_instance_mem_gb
submit_cmd_str += " --workflow-file-subset %s" % (
pipeline)
submit_cmd_str += " --timing-metrics-json %s" % (
timing_json)
submit_cmd_str += " --related-file-subset-uuids %s" % (
",".join(related_uuids))
submit_cmd_str += " --uuid %s" % (
uuid_map[pipeline][donor])
#submit_cmd_str += " --skip-upload"
handle.write(
string.Template("""#!/bin/bash
set -ex
${PREP}
${SUBMIT}
echo $$? > $$0.submitted
#pushd ${SUBMIT_DIR}
#gtupload -v -c ${KEY} -u ./manifest.xml
#ECODE=$$?
#popd
#echo $$ECODE > $$0.uploaded
""").substitute(PREP=prep_cmd_str,
SUBMIT=submit_cmd_str,
SUBMIT_DIR=os.path.join(os.path.abspath(args.workdir), "vcf",
pipeline + "." + donor_tumor + ".dir",
uuid_map[pipeline][donor]),
KEY=args.keyfile))
def run_gen(args):
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"db_snp": "dbsnp_132_b37.leftAligned.vcf",
"centromere": "centromere_hg19.bed",
"reference_genome": "Homo_sapiens_assembly19.fasta",
"cosmic": "b37_cosmic_v54_120711.vcf"
}
if args.ref_download:
syn_sync(syn, REFDATA_PROJECT, docstore, data_mapping.values())
dm = {}
for k, v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = {"uuid": hit}
if args.sample is not None:
sync_doc_dir(os.path.join(os.path.dirname(__file__), "..",
"testexomes"),
docstore,
filter=lambda x: x['donorId'] in args.sample)
else:
sync_doc_dir(
os.path.join(os.path.dirname(__file__), "..", "testexomes"),
docstore)
tumor_uuids = {}
normal_uuids = {}
for id, ent in docstore.filter(sampleType="tumour"):
tumor_uuids[ent['participant_id']] = id
for id, ent in docstore.filter(sampleType="normal"):
normal_uuids[ent['participant_id']] = id
mc3_workflow = GalaxyWorkflow(
ga_file="workflows/Galaxy-Workflow-MC3_Pipeline_Test.ga")
reference_id = None
for a in docstore.filter(name="Homo_sapiens_assembly19.fasta"):
reference_id = a[0]
tasks = TaskGroup()
for donor in tumor_uuids:
if donor in normal_uuids:
print "participant", donor
donor_name = None
for k, v in fake_metadata.items():
if v['participant_id'] == donor:
donor_name = k
workflow_dm = dict(dm)
workflow_dm['tumor_bam'] = {"uuid": tumor_uuids[donor]}
workflow_dm['normal_bam'] = {"uuid": normal_uuids[donor]}
task = GalaxyWorkflowTask(
"workflow_%s" % (donor),
mc3_workflow,
inputs=workflow_dm,
parameters={
"reheader_config": {
"platform":
"Illumina",
"center":
"OHSU",
"reference_genome":
"Homo_sapiens_assembly19.fasta",
"participant_uuid":
fake_metadata[donor_name]['participant_id'],
"disease_code":
fake_metadata[donor_name]['disease'],
"filedate":
datetime.datetime.now().strftime("%Y%m%d"),
"normal_analysis_uuid":
fake_metadata[donor_name]['normal']['uuid'],
"normal_bam_name":
fake_metadata[donor_name]['normal']['file_name'],
"normal_aliquot_uuid":
fake_metadata[donor_name]['normal']['aliquot_id'],
"normal_aliquot_barcode":
fake_metadata[donor_name]['normal']['barcode'],
"tumor_analysis_uuid":
fake_metadata[donor_name]['tumour']['uuid'],
"tumor_bam_name":
fake_metadata[donor_name]['tumour']['file_name'],
"tumor_aliquot_uuid":
fake_metadata[donor_name]['tumour']['aliquot_id'],
"tumor_aliquot_barcode":
fake_metadata[donor_name]['tumour']['barcode'],
}
},
tags=["donor:%s" % (donor)],
)
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id),
"w") as handle:
handle.write(json.dumps(data.to_dict()))
if args.create_service:
service = GalaxyService(docstore=docstore,
galaxy=args.galaxy,
sudo=args.sudo,
tool_data=args.tool_data,
tool_dir=args.tool_dir,
work_dir=args.work_dir,
smp=[["gatk_bqsr", 12], ["gatk_indel", 24],
["MuSE", 8], ["pindel", 8], ["mutect", 8],
["delly", 4], ["gatk_bqsr", 12],
["gatk_indel", 12], ["bwa_mem", 12],
["radia", 8], ['radia_filter', 8]])
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
def run_gen(args):
args = parser.parse_args()
syn = synapseclient.Synapse()
syn.login()
docstore = from_url(args.out_base)
data_mapping = {
"reference_genome": "genome.fa",
"dbsnp": "dbsnp_132_b37.leftAligned.vcf",
"cosmic": "b37_cosmic_v54_120711.vcf",
"gold_indels":
"Mills_and_1000G_gold_standard.indels.hg19.sites.fixed.vcf",
"phase_one_indels": "1000G_phase1.indels.hg19.sites.fixed.vcf",
"centromere": "centromere_hg19.bed"
}
if args.ref_download:
#download reference files from Synapse and populate the document store
for a in syn.chunkedQuery('select * from entity where parentId=="%s"' %
(REFDATA_PROJECT)):
print "found", a['entity.name']
if a['entity.name'] in data_mapping.values(
) or a['entity.name'].replace(".gz", "") in data_mapping.values():
print "loading"
ent = syn.get(a['entity.id'])
id = ent.annotations['uuid'][0]
t = Target(uuid=id)
docstore.create(t)
path = docstore.get_filename(t)
name = ent.name
if 'dataPrep' in ent.annotations:
if ent.annotations['dataPrep'][0] == 'gunzip':
subprocess.check_call("gunzip -c %s > %s" %
(ent.path, path),
shell=True)
name = name.replace(".gz", "")
else:
print "Unknown DataPrep"
else:
shutil.copy(ent.path, path)
docstore.update_from_file(t)
meta = {}
meta['name'] = name
meta['uuid'] = id
if 'dataPrep' in meta:
del meta['dataPrep']
docstore.put(id, meta)
dm = {}
for k, v in data_mapping.items():
hit = None
for a in docstore.filter(name=v):
hit = a[0]
if hit is None:
raise Exception("%s not found" % (v))
dm[k] = {"uuid": hit}
workflow = GalaxyWorkflow(
ga_file="workflows/Galaxy-Workflow-PCAWG_CGHUB.ga")
tasks = TaskGroup()
for ent in synqueue.listAssignments(syn, **config):
#print "'%s'" % (ent['state']), ent['state'] == 'nan', type(ent['state']), type('nan')
if not isinstance(ent['state'], basestring) and isnan(ent['state']):
gnos_endpoint = urlparse(
ent['meta']['Normal_WGS_alignment_GNOS_repos']).netloc
task = GalaxyWorkflowTask(
"workflow_%s" % (ent['id']),
workflow,
inputs=dm,
parameters={
'normal_bam_download': {
"uuid":
ent['meta']['Normal_WGS_alignment_GNOS_analysis_ID'],
"gnos_endpoint":
gnos_endpoint,
"cred_file":
key_map[gnos_endpoint]
},
'tumor_bam_download': {
"uuid":
ent['meta']['Tumour_WGS_alignment_GNOS_analysis_IDs'],
"gnos_endpoint":
gnos_endpoint,
"cred_file":
key_map[gnos_endpoint]
},
'broad_variant_pipeline': {
"broad_ref_dir": "/tool_data/files/refdata",
"sample_id": ent['meta']['Submitter_donor_ID']
}
},
tags=["donor:%s" % (ent['meta']['Submitter_donor_ID'])])
tasks.append(task)
if not os.path.exists("%s.tasks" % (args.out_base)):
os.mkdir("%s.tasks" % (args.out_base))
for data in tasks:
with open("%s.tasks/%s" % (args.out_base, data.task_id),
"w") as handle:
handle.write(json.dumps(data.to_dict()))
state_file = "%s.tasks/%s.state" % (args.out_base, data.task_id)
if os.path.exists(state_file):
os.unlink(state_file)
print "Tasks Created: %s" % (len(tasks))
if args.create_service:
service = GalaxyService(docstore=docstore,
galaxy="bgruening/galaxy-stable",
sudo=args.sudo,
tool_data=os.path.abspath("tool_data"),
tool_dir=os.path.abspath("tools"),
work_dir=args.work_dir,
smp=[["MuSE", 8], ["pindel", 8], ["muTect", 8],
["delly", 4], ["gatk_bqsr", 12],
["gatk_indel", 24], ["bwa_mem", 12],
["broad_variant_pipline", 24]])
with open("%s.service" % (args.out_base), "w") as handle:
s = service.get_config()
if args.scratch:
print "Using scratch", args.scratch
s.set_docstore_config(cache_path=args.scratch, open_perms=True)
s.store(handle)
