def merge_bam_files(bam_files, work_dir, config, batch=0):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Avoids too many open file issues by merging large numbers of files in batches.
"""
max_merge = 500
bam_files.sort()
i = 1
while len(bam_files) > max_merge:
bam_files = [merge_bam_files(xs, work_dir, config, batch + i)
for xs in utils.partition_all(max_merge, bam_files)]
i += 1
if batch > 0:
out_dir = utils.safe_makedir(os.path.join(work_dir, "batchmerge%s" % batch))
else:
out_dir = work_dir
out_file = os.path.join(out_dir, os.path.basename(sorted(bam_files)[0]))
picard = broad.runner_from_config(config)
if len(bam_files) == 1:
if not os.path.exists(out_file):
os.symlink(bam_files[0], out_file)
else:
picard.run_fn("picard_merge", bam_files, out_file)
for b in bam_files:
utils.save_diskspace(b, "BAM merged to %s" % out_file, config)
return out_file
def split_bam_file(bam_file, split_size, out_dir, config):
"""Split a BAM file into paired end fastq splits based on split size.
XXX Need to generalize for non-paired end inputs.
"""
existing = _find_current_bam_split(bam_file, out_dir)
if len(existing) > 0:
return existing
pipe = True
utils.safe_makedir(out_dir)
broad_runner = broad.runner_from_config(config)
out_files = []
def new_handle(num):
out = []
for pair in [1, 2]:
fname = os.path.join(
out_dir,
"{base}_{pair}_{num}.fastq".format(
base=os.path.splitext(os.path.basename(bam_file))[0], pair=pair, num=num
),
)
out += [fname, open(fname, "w")]
return out
with utils.curdir_tmpdir(base_dir=config_utils.get_resources("tmp", config).get("dir")) as tmp_dir:
if pipe:
sort_file = os.path.join(tmp_dir, "%s-sort.bam" % os.path.splitext(os.path.basename(bam_file))[0])
os.mkfifo(sort_file)
broad_runner.run_fn("picard_sort", bam_file, "queryname", sort_file, compression_level=0, pipe=True)
else:
sort_file = os.path.join(out_dir, "%s-sort.bam" % os.path.splitext(os.path.basename(bam_file))[0])
broad_runner.run_fn("picard_sort", bam_file, "queryname", sort_file)
samfile = pysam.Samfile(sort_file, "rb")
i = 0
num = 0
f1, out_handle1, f2, out_handle2 = new_handle(num)
out_files.append([f1, f2, None])
for x1, x2 in utils.partition_all(2, samfile):
x1_seq, x1_qual = _get_seq_qual(x1)
out_handle1.write("@%s/1\n%s\n+\n%s\n" % (i, x1_seq, x1_qual))
x2_seq, x2_qual = _get_seq_qual(x2)
out_handle2.write("@%s/2\n%s\n+\n%s\n" % (i, x2_seq, x2_qual))
i += 1
if i % split_size == 0:
num += 1
out_handle1.close()
out_handle2.close()
f1, out_handle1, f2, out_handle2 = new_handle(num)
out_files.append([f1, f2, num])
out_handle1.close()
out_handle2.close()
samfile.close()
if pipe:
os.unlink(sort_file)
else:
utils.save_diskspace(sort_file, "Split to {}".format(out_files[0][0]), config)
return out_files
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
if len(bam_files) == 1 and bam.bam_already_sorted(bam_files[0], config,
"coordinate"):
shutil.copy(bam_files[0], out_file)
else:
if out_file is None:
out_file = os.path.join(work_dir,
os.path.basename(sorted(bam_files)[0]))
if batch is not None:
base, ext = os.path.splitext(out_file)
out_file = "%s-b%s%s" % (base, batch, ext)
if not utils.file_exists(out_file):
sambamba = config_utils.get_program("sambamba", config)
samtools = config_utils.get_program("samtools", config)
resources = config_utils.get_resources("samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(resources.get("memory", "1G"),
2, "decrease").upper()
# sambamba opens 4 handles per file, so try to guess a reasonable batch size
batch_size = (system.open_file_limit() // 4) - 100
if len(bam_files) > batch_size:
bam_files = [
merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(
utils.partition_all(batch_size, bam_files))
]
with tx_tmpdir(config) as tmpdir:
with utils.chdir(tmpdir):
with file_transaction(config, out_file) as tx_out_file:
with file_transaction(
config,
"%s.list" % os.path.splitext(out_file)[0]
) as tx_bam_file_list:
with open(tx_bam_file_list, "w") as out_handle:
for f in sorted(bam_files):
out_handle.write("%s\n" % f)
if bam.bam_already_sorted(bam_files[0], config,
"coordinate"):
cmd = _sambamba_merge(bam_files)
else:
assert config.get("mark_duplicates", True)
cmd = _biobambam_merge_dedup()
do.run(
cmd.format(**locals()),
"Merge bam files to %s" %
os.path.basename(out_file), None)
# Ensure timestamps are up to date on output file and index
# Works around issues on systems with inconsistent times
for ext in ["", ".bai"]:
if os.path.exists(out_file + ext):
subprocess.check_call(["touch", out_file + ext])
for b in bam_files:
utils.save_diskspace(b, "BAM merged to %s" % out_file, config)
bam.index(out_file, config)
return out_file
def split_bam_file(bam_file, split_size, out_dir, config):
"""Split a BAM file into paired end fastq splits based on split size.
XXX Need to generalize for non-paired end inputs.
"""
existing = _find_current_bam_split(bam_file, out_dir)
if len(existing) > 0:
return existing
pipe = True
utils.safe_makedir(out_dir)
broad_runner = broad.runner_from_config(config)
out_files = []
def new_handle(num):
out = []
for pair in [1, 2]:
fname = os.path.join(out_dir, "{base}_{pair}_{num}.fastq".format(
base=os.path.splitext(os.path.basename(bam_file))[0], pair=pair, num=num))
out += [fname, open(fname, "w")]
return out
with utils.curdir_tmpdir(base_dir=config_utils.get_resources("tmp", config).get("dir")) as tmp_dir:
if pipe:
sort_file = os.path.join(tmp_dir, "%s-sort.bam" %
os.path.splitext(os.path.basename(bam_file))[0])
os.mkfifo(sort_file)
broad_runner.run_fn("picard_sort", bam_file, "queryname", sort_file,
compression_level=0, pipe=True)
else:
sort_file = os.path.join(out_dir, "%s-sort.bam" %
os.path.splitext(os.path.basename(bam_file))[0])
broad_runner.run_fn("picard_sort", bam_file, "queryname", sort_file)
samfile = pysam.Samfile(sort_file, "rb")
i = 0
num = 0
f1, out_handle1, f2, out_handle2 = new_handle(num)
out_files.append([f1, f2, None])
for x1, x2 in utils.partition_all(2, samfile):
x1_seq, x1_qual = _get_seq_qual(x1)
out_handle1.write("@%s/1\n%s\n+\n%s\n" % (i, x1_seq, x1_qual))
x2_seq, x2_qual = _get_seq_qual(x2)
out_handle2.write("@%s/2\n%s\n+\n%s\n" % (i, x2_seq, x2_qual))
i += 1
if i % split_size == 0:
num += 1
out_handle1.close()
out_handle2.close()
f1, out_handle1, f2, out_handle2 = new_handle(num)
out_files.append([f1, f2, num])
out_handle1.close()
out_handle2.close()
samfile.close()
if pipe:
os.unlink(sort_file)
else:
utils.save_diskspace(sort_file, "Split to {}".format(out_files[0][0]), config)
return out_files
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
out_file = _merge_outfile_fname(out_file, bam_files, work_dir, batch)
if not utils.file_exists(out_file):
if len(bam_files) == 1 and bam.bam_already_sorted(
bam_files[0], config, "coordinate"):
with file_transaction(config, out_file) as tx_out_file:
_create_merge_filelist(bam_files, tx_out_file, config)
shutil.copy(bam_files[0], tx_out_file)
samtools = config_utils.get_program("samtools", config)
do.run('{} quickcheck -v {}'.format(samtools, out_file),
"Check for valid merged BAM after transfer")
else:
# sambamba opens 4 handles per file, so try to guess a reasonable batch size
batch_size = (system.open_file_limit() // 4) - 100
if len(bam_files) > batch_size:
bam_files = [
merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(
utils.partition_all(batch_size, bam_files))
]
with tx_tmpdir(config) as tmpdir:
with utils.chdir(tmpdir):
with file_transaction(config, out_file) as tx_out_file:
tx_bam_file_list = _create_merge_filelist(
bam_files, tx_out_file, config)
sambamba = config_utils.get_program("sambamba", config)
samtools = config_utils.get_program("samtools", config)
resources = config_utils.get_resources(
"samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(
resources.get("memory", "1G"), 2,
"decrease").upper()
if bam.bam_already_sorted(bam_files[0], config,
"coordinate"):
cmd = _sambamba_merge(bam_files)
else:
assert config.get("mark_duplicates", True)
cmd = _biobambam_merge_dedup()
do.run(
cmd.format(**locals()), "Merge bam files to %s" %
os.path.basename(out_file), None)
do.run(
'{} quickcheck -v {}'.format(
samtools, tx_out_file),
"Check for valid merged BAM")
do.run('{} quickcheck -v {}'.format(samtools, out_file),
"Check for valid merged BAM after transfer")
_finalize_merge(out_file, bam_files, config)
bam.index(out_file, config)
return out_file
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
if len(bam_files) == 1:
return bam_files[0]
else:
if out_file is None:
out_file = os.path.join(work_dir,
os.path.basename(sorted(bam_files)[0]))
if batch is not None:
base, ext = os.path.splitext(out_file)
out_file = "%s-b%s%s" % (base, batch, ext)
if not utils.file_exists(out_file) or not utils.file_exists(out_file +
".bai"):
bamtools = config_utils.get_program("bamtools", config)
samtools = config_utils.get_program("samtools", config)
resources = config_utils.get_resources("samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(resources.get("memory", "1G"),
2, "decrease").upper()
batch_size = system.open_file_limit() - 100
if len(bam_files) > batch_size:
bam_files = [
merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(
utils.partition_all(batch_size, bam_files))
]
with utils.curdir_tmpdir({"config": config}) as tmpdir:
with utils.chdir(tmpdir):
merge_cl = _bamtools_merge(bam_files)
with file_transaction(out_file) as tx_out_file:
with file_transaction("%s.list" %
os.path.splitext(out_file)[0]
) as tx_bam_file_list:
tx_out_prefix = os.path.splitext(tx_out_file)[0]
with open(tx_bam_file_list, "w") as out_handle:
for f in sorted(bam_files):
out_handle.write("%s\n" % f)
cmd = (
merge_cl + " | "
"{samtools} sort -@ {num_cores} -m {max_mem} - {tx_out_prefix}"
)
do.run(
cmd.format(**locals()),
"Merge bam files to %s" %
os.path.basename(out_file), None)
for b in bam_files:
utils.save_diskspace(b, "BAM merged to %s" % out_file, config)
bam.index(out_file, config)
return out_file
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
if len(bam_files) == 1:
bam.index(bam_files[0], config)
return bam_files[0]
else:
if out_file is None:
out_file = os.path.join(work_dir, os.path.basename(sorted(bam_files)[0]))
if batch is not None:
base, ext = os.path.splitext(out_file)
out_file = "%s-b%s%s" % (base, batch, ext)
if not utils.file_exists(out_file):
sambamba = config_utils.get_program("sambamba", config)
samtools = config_utils.get_program("samtools", config)
samblaster = config_utils.get_program("samblaster", config)
resources = config_utils.get_resources("samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(resources.get("memory", "1G"),
2, "decrease").upper()
# sambamba opens 4 handles per file, so try to guess a reasonable batch size
batch_size = (system.open_file_limit() // 4) - 100
if len(bam_files) > batch_size:
bam_files = [merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(utils.partition_all(batch_size, bam_files))]
with tx_tmpdir(config) as tmpdir:
with utils.chdir(tmpdir):
with file_transaction(config, out_file) as tx_out_file:
with file_transaction(config, "%s.list" % os.path.splitext(out_file)[0]) as tx_bam_file_list:
with open(tx_bam_file_list, "w") as out_handle:
for f in sorted(bam_files):
out_handle.write("%s\n" % f)
if bam.bam_already_sorted(bam_files[0], config, "coordinate"):
cmd = _sambamba_merge(bam_files)
else:
assert config.get("mark_duplicates", True)
cmd = _biobambam_merge_dedup()
do.run(cmd.format(**locals()), "Merge bam files to %s" % os.path.basename(out_file),
None)
# Ensure timestamps are up to date on output file and index
# Works around issues on systems with inconsistent times
for ext in ["", ".bai"]:
if os.path.exists(out_file + ext):
subprocess.check_call(["touch", out_file + ext])
for b in bam_files:
utils.save_diskspace(b, "BAM merged to %s" % out_file, config)
bam.index(out_file, config)
return out_file
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
out_file = _merge_outfile_fname(out_file, bam_files, work_dir, batch)
if not utils.file_exists(out_file):
if len(bam_files) == 1 and bam.bam_already_sorted(bam_files[0], config, "coordinate"):
with file_transaction(config, out_file) as tx_out_file:
_create_merge_filelist(bam_files, tx_out_file, config)
shutil.copy(bam_files[0], tx_out_file)
samtools = config_utils.get_program("samtools", config)
do.run('{} quickcheck -v {}'.format(samtools, out_file),
"Check for valid merged BAM after transfer")
else:
# sambamba opens 4 handles per file, so try to guess a reasonable batch size
batch_size = (system.open_file_limit() // 4) - 100
if len(bam_files) > batch_size:
bam_files = [merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(utils.partition_all(batch_size, bam_files))]
with tx_tmpdir(config) as tmpdir:
with utils.chdir(tmpdir):
with file_transaction(config, out_file) as tx_out_file:
tx_bam_file_list = _create_merge_filelist(bam_files, tx_out_file, config)
sambamba = config_utils.get_program("sambamba", config)
samtools = config_utils.get_program("samtools", config)
resources = config_utils.get_resources("samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(resources.get("memory", "1G"),
2, "decrease").upper()
if bam.bam_already_sorted(bam_files[0], config, "coordinate"):
cmd = _sambamba_merge(bam_files)
else:
# Aim for 3.5Gb/core memory for BAM merging
num_cores = config_utils.adjust_cores_to_mb_target(
3500, resources.get("memory", "2G"), num_cores)
assert config.get("mark_duplicates", True)
cmd = _biobambam_merge_dedup()
do.run(cmd.format(**locals()), "Merge bam files to %s" % os.path.basename(out_file),
None)
do.run('{} quickcheck -v {}'.format(samtools, tx_out_file),
"Check for valid merged BAM")
do.run('{} quickcheck -v {}'.format(samtools, out_file),
"Check for valid merged BAM after transfer")
_finalize_merge(out_file, bam_files, config)
bam.index(out_file, config)
return out_file
def run(align_bams, items, ref_file, assoc_files, region, out_file):
"""Run platypus variant calling, germline whole genome or exome.
"""
assert out_file.endswith(".vcf.gz")
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
for align_bam in align_bams:
bam.index(align_bam, items[0]["config"])
cmd = [
"platypus", "callVariants",
"--regions=%s" % _subset_regions(region, out_file, items),
"--bamFiles=%s" % ",".join(align_bams),
"--refFile=%s" % dd.get_ref_file(items[0]), "--output=-",
"--logFileName", "/dev/null", "--verbosity=1"
]
resources = config_utils.get_resources("platypus",
items[0]["config"])
if resources.get("options"):
# normalize options so we can set defaults without overwriting user specified
for opt in resources["options"]:
if "=" in opt:
key, val = opt.split("=")
cmd.extend([key, val])
else:
cmd.append(opt)
# Adjust default filter thresholds to achieve similar sensitivity/specificity to other callers
tuned_opts = [
"--hapScoreThreshold", "10", "--scThreshold", "0.99",
"--filteredReadsFrac", "0.9", "--rmsmqThreshold", "20",
"--qdThreshold", "0", "--abThreshold", "0.0001",
"--minVarFreq", "0.0", "--assemble", "1"
]
for okey, oval in utils.partition_all(2, tuned_opts):
if okey not in cmd:
cmd.extend([okey, oval])
# Avoid filtering duplicates on high depth targeted regions where we don't mark duplicates
if any(not tz.get_in(["config", "algorithm", "mark_duplicates"],
data, True) for data in items):
cmd += ["--filterDuplicates=0"]
post_process_cmd = (
" | %s | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | "
"vcfstreamsort | bgzip -c > %s" %
(vcfutils.fix_ambiguous_cl(), tx_out_file))
do.run(" ".join(cmd) + post_process_cmd,
"platypus variant calling")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def merge_bam_files(bam_files, work_dir, config, out_file=None, batch=None):
"""Merge multiple BAM files from a sample into a single BAM for processing.
Checks system open file limit and merges in batches if necessary to avoid
file handle limits.
"""
if len(bam_files) == 1:
return bam_files[0]
else:
if out_file is None:
out_file = os.path.join(work_dir, os.path.basename(sorted(bam_files)[0]))
if batch is not None:
base, ext = os.path.splitext(out_file)
out_file = "%s-b%s%s" % (base, batch, ext)
if not utils.file_exists(out_file) or not utils.file_exists(out_file + ".bai"):
bamtools = config_utils.get_program("bamtools", config)
samtools = config_utils.get_program("samtools", config)
resources = config_utils.get_resources("samtools", config)
num_cores = config["algorithm"].get("num_cores", 1)
max_mem = config_utils.adjust_memory(resources.get("memory", "1G"),
2, "decrease")
batch_size = system.open_file_limit() - 100
if len(bam_files) > batch_size:
bam_files = [merge_bam_files(xs, work_dir, config, out_file, i)
for i, xs in enumerate(utils.partition_all(batch_size, bam_files))]
with utils.curdir_tmpdir({"config": config}) as tmpdir:
with utils.chdir(tmpdir):
merge_cl = _bamtools_merge(bam_files)
with file_transaction(out_file) as tx_out_file:
with file_transaction("%s.list" % os.path.splitext(out_file)[0]) as tx_bam_file_list:
tx_out_prefix = os.path.splitext(tx_out_file)[0]
with open(tx_bam_file_list, "w") as out_handle:
for f in sorted(bam_files):
out_handle.write("%s\n" % f)
cmd = (merge_cl + " | "
"{samtools} sort -@ {num_cores} -m {max_mem} - {tx_out_prefix}")
do.run(cmd.format(**locals()), "Merge bam files to %s" % os.path.basename(out_file),
None)
for b in bam_files:
utils.save_diskspace(b, "BAM merged to %s" % out_file, config)
bam.index(out_file, config)
return out_file
def _combine_variants(in_vcfs, out_file, ref_file, config):
"""Combine variant files, batching to avoid problematic large commandlines.
"""
max_batch = 500
if len(in_vcfs) > max_batch:
new_vcfs = []
for i, batch_vcfs in enumerate(partition_all(max_batch, in_vcfs)):
path, fname = os.path.split(out_file)
batch_path = safe_makedir(os.path.join(path, "batch"))
base, ext = os.path.splitext(fname)
cur_out = os.path.join(batch_path, "{0}-batch{1}{2}".format(base, i, ext))
for x in batch_vcfs:
with open(x) as in_handle:
if not in_handle.readline().startswith("##fileformat=VCFv4"):
raise ValueError("Unexpected VCF file: %s" % x)
combine_variant_files(batch_vcfs, cur_out, ref_file, config)
new_vcfs.append(cur_out)
in_vcfs = new_vcfs
assert len(in_vcfs) <= max_batch
combine_variant_files(in_vcfs, out_file, ref_file, config)
def _combine_variants(in_vcfs, out_file, ref_file, config):
"""Combine variant files, batching to avoid problematic large commandlines.
"""
max_batch = 500
if len(in_vcfs) > max_batch:
new_vcfs = []
for i, batch_vcfs in enumerate(partition_all(max_batch, in_vcfs)):
path, fname = os.path.split(out_file)
batch_path = safe_makedir(os.path.join(path, "batch"))
base, ext = os.path.splitext(fname)
cur_out = os.path.join(batch_path,
"{0}-batch{1}{2}".format(base, i, ext))
for x in batch_vcfs:
with open(x) as in_handle:
if not in_handle.readline().startswith(
"##fileformat=VCFv4"):
raise ValueError("Unexpected VCF file: %s" % x)
combine_variant_files(batch_vcfs, cur_out, ref_file, config)
new_vcfs.append(cur_out)
in_vcfs = new_vcfs
assert len(in_vcfs) <= max_batch
combine_variant_files(in_vcfs, out_file, ref_file, config)
def run(align_bams, items, ref_file, assoc_files, region, out_file):
"""Run platypus variant calling, germline whole genome or exome.
"""
assert out_file.endswith(".vcf.gz")
if not utils.file_exists(out_file):
with file_transaction(items[0], out_file) as tx_out_file:
for align_bam in align_bams:
bam.index(align_bam, items[0]["config"])
cmd = ["platypus", "callVariants", "--regions=%s" % _subset_regions(region, out_file, items),
"--bamFiles=%s" % ",".join(align_bams),
"--refFile=%s" % dd.get_ref_file(items[0]), "--output=-",
"--logFileName", "/dev/null", "--verbosity=1"]
resources = config_utils.get_resources("platypus", items[0]["config"])
if resources.get("options"):
# normalize options so we can set defaults without overwriting user specified
for opt in resources["options"]:
if "=" in opt:
key, val = opt.split("=")
cmd.extend([key, val])
else:
cmd.append(opt)
# Adjust default filter thresholds to achieve similar sensitivity/specificity to other callers
tuned_opts = ["--hapScoreThreshold", "10", "--scThreshold", "0.99", "--filteredReadsFrac", "0.9",
"--rmsmqThreshold", "20", "--qdThreshold", "0", "--abThreshold", "0.0001",
"--minVarFreq", "0.0", "--assemble", "1"]
for okey, oval in utils.partition_all(2, tuned_opts):
if okey not in cmd:
cmd.extend([okey, oval])
# Avoid filtering duplicates on high depth targeted regions where we don't mark duplicates
if any(not tz.get_in(["config", "algorithm", "mark_duplicates"], data, True)
for data in items):
cmd += ["--filterDuplicates=0"]
post_process_cmd = (" | %s | vcfallelicprimitives -t DECOMPOSED --keep-geno | vcffixup - | "
"vcfstreamsort | bgzip -c > %s" % (vcfutils.fix_ambiguous_cl(), tx_out_file))
do.run(" ".join(cmd) + post_process_cmd, "platypus variant calling")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file