def samples_to_records(samples, default_keys=None):
"""Convert samples into output CWL records.
"""
from bcbio.pipeline import run_info
RECORD_CONVERT_TO_LIST = set([
"config__algorithm__tools_on", "config__algorithm__tools_off",
"reference__genome_context"
])
all_keys = _get_all_cwlkeys(samples, default_keys)
out = []
for data in samples:
for raw_key in sorted(list(all_keys)):
key = raw_key.split("__")
if tz.get_in(key, data) is None:
data = tz.update_in(data, key, lambda x: None)
if raw_key not in data["cwl_keys"]:
data["cwl_keys"].append(raw_key)
if raw_key in RECORD_CONVERT_TO_LIST:
val = tz.get_in(key, data)
if not val: val = []
elif not isinstance(val, (list, tuple)): val = [val]
data = tz.update_in(data, key, lambda x: val)
# Booleans are problematic for CWL serialization, convert into string representation
if isinstance(tz.get_in(key, data), bool):
data = tz.update_in(data, key,
lambda x: str(tz.get_in(key, data)))
data["metadata"] = run_info.add_metadata_defaults(
data.get("metadata", {}))
out.append(data)
return out
def samples_to_records(samples, default_keys=None):
"""Convert samples into output CWL records.
"""
from bcbio.pipeline import run_info
RECORD_CONVERT_TO_LIST = set(["config__algorithm__tools_on", "config__algorithm__tools_off",
"reference__genome_context"])
all_keys = _get_all_cwlkeys(samples, default_keys)
out = []
for data in samples:
for raw_key in sorted(list(all_keys)):
key = raw_key.split("__")
if tz.get_in(key, data) is None:
data = tz.update_in(data, key, lambda x: None)
if raw_key not in data["cwl_keys"]:
data["cwl_keys"].append(raw_key)
if raw_key in RECORD_CONVERT_TO_LIST:
val = tz.get_in(key, data)
if not val: val = []
elif not isinstance(val, (list, tuple)): val = [val]
data = tz.update_in(data, key, lambda x: val)
# Booleans are problematic for CWL serialization, convert into string representation
if isinstance(tz.get_in(key, data), bool):
data = tz.update_in(data, key, lambda x: str(tz.get_in(key, data)))
data["metadata"] = run_info.add_metadata_defaults(data.get("metadata", {}))
out.append(data)
return out
def samples_to_records(samples):
"""Convert samples into output CWL records.
"""
from bcbio.pipeline import run_info
RECORD_CONVERT_TO_LIST = set([
"config__algorithm__tools_on", "config__algorithm__tools_off",
"config__algorithm__svcaller"
])
all_keys = _get_all_cwlkeys(samples)
out = []
for data in samples:
for raw_key in sorted(list(all_keys)):
key = raw_key.split("__")
if tz.get_in(key, data) is None:
data = tz.update_in(data, key, lambda x: None)
data["cwl_keys"].append(raw_key)
if raw_key in RECORD_CONVERT_TO_LIST:
val = tz.get_in(key, data)
if not val: val = []
elif not isinstance(val, (list, tuple)): val = [val]
data = tz.update_in(data, key, lambda x: val)
data["metadata"] = run_info.add_metadata_defaults(
data.get("metadata", {}))
out.append(data)
return out
def _samples_to_records(samples):
"""Convert samples into output CWL records.
"""
from bcbio.pipeline import run_info
RECORD_CONVERT_TO_LIST = set(["config__algorithm__tools_on", "config__algorithm__tools_off"])
all_keys = _get_all_cwlkeys(samples)
out = []
for data in samples:
for raw_key in sorted(list(all_keys)):
key = raw_key.split("__")
if tz.get_in(key, data) is None:
data = tz.update_in(data, key, lambda x: None)
data["cwl_keys"].append(raw_key)
if raw_key in RECORD_CONVERT_TO_LIST:
val = tz.get_in(key, data)
if not val: val = []
elif not isinstance(val, (list, tuple)): val = [val]
data = tz.update_in(data, key, lambda x: val)
data["metadata"] = run_info.add_metadata_defaults(data.get("metadata", {}))
out.append(data)
return out
def batch_for_variantcall(samples):
"""Prepare a set of samples for parallel variant calling.
CWL input target that groups samples into batches and variant callers
for parallel processing.
"""
from bcbio.pipeline import run_info
convert_to_list = set(["config__algorithm__tools_on", "config__algorithm__tools_off"])
default_keys = set(["metadata__batch", "config__algorithm__validate",
"config__algorithm__validate_regions"])
to_process, extras = _dup_samples_by_variantcaller(samples, require_bam=False)
batch_groups = collections.defaultdict(list)
to_process = [utils.to_single_data(x) for x in to_process]
all_keys = set([])
for data in to_process:
all_keys.update(set(data["cwl_keys"]))
all_keys.update(default_keys)
for data in to_process:
for raw_key in sorted(list(all_keys)):
key = raw_key.split("__")
if tz.get_in(key, data) is None:
data = tz.update_in(data, key, lambda x: None)
data["cwl_keys"].append(raw_key)
if raw_key in convert_to_list:
val = tz.get_in(key, data)
if not val: val = []
elif not isinstance(val, (list, tuple)): val = [val]
data = tz.update_in(data, key, lambda x: val)
vc = get_variantcaller(data, require_bam=False)
data["metadata"] = run_info.add_metadata_defaults(data.get("metadata", {}))
batches = dd.get_batches(data) or dd.get_sample_name(data)
if not isinstance(batches, (list, tuple)):
batches = [batches]
for b in batches:
batch_groups[(b, vc)].append(utils.deepish_copy(data))
return list(batch_groups.values()) + extras
