def _reallocate(in_file):
tool = do.find_cmd("reallocate.pl")
cmd = "perl {tool} -i {in_file} 5000 1000 b"
out_file = in_file + ".weighted-5000-1000-b"
if not utils.file_exists(out_file):
do.run(cmd.format(**locals()), "reallocate")
return out_file
def _bedpe_to_vcf(bedpe_file, sconfig_file, items):
"""Convert BEDPE output into a VCF file.
"""
tovcf_script = do.find_cmd("bedpeToVcf")
if tovcf_script:
out_file = "%s.vcf.gz" % utils.splitext_plus(bedpe_file)[0]
out_nogzip = out_file.replace(".vcf.gz", ".vcf")
raw_file = "%s-raw.vcf" % utils.splitext_plus(bedpe_file)[0]
if not utils.file_exists(out_file):
if not utils.file_exists(raw_file):
with file_transaction(raw_file) as tx_raw_file:
ref_file = tz.get_in(["reference", "fasta", "base"], items[0])
cmd = [
sys.executable,
tovcf_script,
"-c",
sconfig_file,
"-f",
ref_file,
"-b",
bedpe_file,
"-o",
tx_raw_file,
]
do.run(cmd, "Convert lumpy bedpe output to VCF")
prep_file = vcfutils.sort_by_ref(raw_file, items[0])
if not utils.file_exists(out_nogzip):
utils.symlink_plus(prep_file, out_nogzip)
out_file = vcfutils.bgzip_and_index(out_nogzip, items[0]["config"])
return out_file
def _clean(in_file):
tool = do.find_cmd("TBr2_duster.pl")
cmd = "perl {tool} -i {in_file}"
out_file = in_file + ".no-dust"
if not utils.file_exists(out_file):
do.run(cmd.format(**locals()), "duster")
return out_file
def _call_vcf(in_bam, sample, workdir, reference, config):
"""
recalibration from BisSNP tool
"""
bissnp = do.find_cmd("bissnp")
basename = sample
num_cores = config['algorithm'].get('cores', 1)
memory = config['algorithm'].get('memory', 4)
jvm_opts = "-Xms750m -Xmx%sg" % memory
cmd = ("{bissnp} {jvm_opts} -R {reference} -I {in_bam} "
"-T BisulfiteGenotyper "
"-vfn1 {tx_out} "
"-vfn2 {out_vfn2} "
"-stand_call_conf 20 "
"-stand_emit_conf 0 "
"-mmq 0 "
"-mbq 0 "
"-nt {num_cores}")
with chdir(workdir):
out_vfn1 = op.join(workdir, sample + ".rawcpg.vcf")
out_vfn2 = op.join(workdir, sample + ".rawsnp.vcf")
if not file_exists(out_vfn1):
with file_transaction(out_vfn1) as tx_out:
log.logger.debug(cmd.format(**locals()))
do.run(cmd.format(**locals()), "BisSNP writerecal in %s" % in_bam)
return out_vfn1, out_vfn2
def _protac(in_file, reference):
tool = do.find_cmd("proTRAC.pl")
cmd = "perl {tool} -genome {reference} -map {in_file} -nh -nr -rpm -distr 1-90 -pimax 32"
out_file = "protac"
if not utils.file_exists(out_file):
do.run(cmd.format(**locals()), "protac")
open(out_file, 'w').close()
return out_file
def _mapper(in_file, reference):
tool = do.find_cmd("sRNAmapper.pl")
cmd = "perl {tool} -i {in_file} -g {reference} -a best -o {tx_out}"
out_file = "hits.eland"
if not utils.file_exists(out_file):
with file_transaction(out_file) as tx_out:
do.run(cmd.format(**locals()), "mapper")
return out_file
def _collapse(in_file):
tool = do.find_cmd("TBr2_collapse.pl")
cmd = "perl {tool} -i {in_file} -o {tx_out}"
basename = splitext_plus(op.basename(in_file))[0]
out_file = basename + "_collapse.fastq"
if not utils.file_exists(out_file):
with file_transaction(out_file) as tx_out:
do.run(cmd.format(**locals()), "collapse")
return out_file
def _align(in_fastq, sample, workdir, genome_index, is_directional, bowtie2,
reference, config):
""" align with bismark. this is actually not used. the align is in ngsalign.bismark.align """
bismark = do.find_cmd("bismark")
resources = config_utils.get_resources("bismark")
num_cores = 1
if resources and resources.get("bismark_threads"):
num_cores = resources.get("bismark_threads")
else:
num_cores = max(int(config['algorithm'].get('cores', 1) / 2), 1)
bowtie_threads = 1
if resources and resources.get("bowtie_threads"):
bowtie_threads = resources.get("bowtie_threads")
basename = sample
if is_directional:
is_directional = ""
else:
is_directional = "--non_directional"
cmd = "{bismark} -n 1 -o {tx_dir} --basename {sample} --unmapped {is_directional} {genome_index} {in_fastq}"
if bowtie2:
cmd = "{bismark} --bowtie2 --parallel {num_cores} -p {bowtie_threads} -o {tx_dir} --basename {sample} --unmapped {is_directional} {genome_index} {in_fastq}"
out_dir = op.join(workdir, sample)
out_bam = op.join(out_dir, sample + ".bam")
with chdir(workdir):
if not file_exists(out_bam):
with tx_tmpdir() as tx_dir:
cmd = cmd.format(**locals())
log.logger.debug(cmd)
do.run(cmd, "bismark in %s" % in_fastq)
shutil.move(tx_dir, out_dir)
broad_runner = broad.runner_from_config(config)
# out_bam, _ = broad_runner.run_fn("picard_formatconverter", out_sam)
names = {
'rg': in_fastq,
'library': 'BS_LIB',
'pl': 'Illumina',
'pu': 'R1',
'sm': in_fastq,
'sample': sample
}
out_fix_bam = broad_runner.run_fn("picard_fix_rgs", out_bam, names)
order_bam = splitext_plus(out_fix_bam)[0] + "_order.bam"
broad_runner.run_fn("picard_reorder", out_fix_bam, reference,
order_bam)
index(order_bam, config)
if bowtie2:
order_bam = _set_quality(order_bam)
index(order_bam, config)
return order_bam
def _bedpe_to_vcf(bedpe_file, sconfig_file, items):
"""Convert BEDPE output into a VCF file.
"""
tovcf_script = do.find_cmd("bedpeToVcf")
if tovcf_script:
out_file = "%s.vcf" % utils.splitext_plus(bedpe_file)[0]
if not utils.file_exists(out_file) and not utils.file_exists(out_file + ".gz"):
with file_transaction(out_file) as tx_out_file:
ref_file = tz.get_in(["reference", "fasta", "base"], items[0])
cmd = [sys.executable, tovcf_script, "-c", sconfig_file, "-f", ref_file,
"-b", bedpe_file, "-o", tx_out_file]
do.run(cmd, "Convert lumpy bedpe output to VCF")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def _run_report(bam_in, sample, biasm_file, workdir, config):
"""
Run bismark2report command
"""
bismark = do.find_cmd("bismark2report")
bam_report = op.join(op.dirname(bam_in), sample) + '_SE_report.txt'
cmd = "{bismark} --alignment_report {bam_report} -o {tx_out} --mbias_report {biasm_file}"
out_dir = op.join(workdir, sample)
out_file = op.join(out_dir, sample + '.html')
with chdir(out_dir):
if not file_exists(out_file):
with file_transaction(out_file) as tx_out:
do.run(cmd.format(**locals()), "bismarkr2report in %s" % bam_in)
return out_dir
def _bedpe_to_vcf(bedpe_file, sconfig_file, items):
"""Convert BEDPE output into a VCF file.
"""
tovcf_script = do.find_cmd("bedpeToVcf")
if tovcf_script:
out_file = "%s.vcf" % utils.splitext_plus(bedpe_file)[0]
if not utils.file_exists(out_file) and not utils.file_exists(out_file +
".gz"):
with file_transaction(out_file) as tx_out_file:
ref_file = tz.get_in(["reference", "fasta", "base"], items[0])
cmd = [
sys.executable, tovcf_script, "-c", sconfig_file, "-f",
ref_file, "-b", bedpe_file, "-o", tx_out_file
]
do.run(cmd, "Convert lumpy bedpe output to VCF")
out_file = vcfutils.bgzip_and_index(out_file, items[0]["config"])
return out_file
def _run_meth_extractor(bam_in, sample, workdir, config):
"""
Run bismark_methylation_extractor command
"""
bismark = do.find_cmd("bismark_methylation_extractor")
cores = config['algorithm'].get('cores', 1)
memory = config['algorithm'].get('mem', 5)
cmd = "{bismark} --no_overlap --comprehensive --multicore {cores} --buffer_size {memory}G --bedGraph --counts --gzip {bam_in}"
out_dir = op.join(workdir, sample)
mbias_file = op.join(out_dir, op.basename(splitext_plus(bam_in)[0]) + '.M-bias.txt')
if not file_exists(mbias_file):
with tx_tmpdir() as tx_dir:
with chdir(tx_dir):
do.run(cmd.format(**locals()), "bismark_methylation_extractor in %s" % bam_in)
shutil.move(tx_dir, out_dir)
assert op.exists(mbias_file), "mbias report doesn't exists:%s" % mbias_file
return mbias_file
def _bedpe_to_vcf(bedpe_file, sconfig_file, items):
"""Convert BEDPE output into a VCF file.
"""
tovcf_script = do.find_cmd("bedpeToVcf")
if tovcf_script:
out_file = "%s.vcf.gz" % utils.splitext_plus(bedpe_file)[0]
out_nogzip = out_file.replace(".vcf.gz", ".vcf")
raw_file = "%s-raw.vcf" % utils.splitext_plus(bedpe_file)[0]
if not utils.file_exists(out_file):
if not utils.file_exists(raw_file):
with file_transaction(items[0], raw_file) as tx_raw_file:
cmd = [sys.executable, tovcf_script, "-c", sconfig_file, "-f", dd.get_ref_file(items[0]),
"-t", "LUMPY", "-b", bedpe_file, "-o", tx_raw_file]
do.run(cmd, "Convert lumpy bedpe output to VCF")
prep_file = vcfutils.sort_by_ref(raw_file, items[0])
if not utils.file_exists(out_nogzip):
utils.symlink_plus(prep_file, out_nogzip)
out_file = vcfutils.bgzip_and_index(out_nogzip, items[0]["config"])
return out_file
def _recal_BQ_score(in_bam, sample, workdir, counts_file, reference, config):
"""
recalibration from BisSNP tool
"""
bissnp = do.find_cmd("bissnp")
basename = sample
num_cores = config['algorithm'].get('cores', 1)
memory = config['algorithm'].get('memory', 4)
jvm_opts = "-Xms750m -Xmx%sg" % memory
cmd = ("{bissnp} {jvm_opts} -R {reference} -I {in_bam} "
"-T BisulfiteTableRecalibration "
"-recalFile {counts_file} "
"-o {tx_out} "
"-maxQ 60 ")
with chdir(workdir):
out_recal = op.join(workdir, sample + "_recal1.bam")
if not file_exists(out_recal):
with file_transaction(out_recal) as tx_out:
log.logger.debug(cmd.format(**locals()))
do.run(cmd.format(**locals()), "BisSNP writerecal in %s" % in_bam)
index(out_recal, config)
return out_recal
def _align(in_fastq, sample, workdir, genome_index, is_directional, bowtie2, reference, config):
"""
align with bismark
"""
bismark = do.find_cmd("bismark")
num_cores = max(int(config['algorithm'].get('cores', 1) / 2), 1)
basename = sample
if is_directional:
is_directional = ""
else:
is_directional = "--non_directional"
cmd = "{bismark} -n 1 -o {tx_dir} --basename {sample} --unmapped {is_directional} {genome_index} {in_fastq}"
if bowtie2:
cmd = "{bismark} --bowtie2 -p {num_cores} -n 1 -o {tx_dir} --basename {sample} --unmapped {is_directional} {genome_index} {in_fastq}"
out_dir = op.join(workdir, sample)
out_bam = op.join(out_dir, sample + ".bam")
with chdir(workdir):
if not file_exists(out_bam):
with tx_tmpdir() as tx_dir:
cmd = cmd.format(**locals())
log.logger.debug(cmd)
do.run(cmd, "bismark in %s" % in_fastq)
shutil.move(tx_dir, out_dir)
broad_runner = broad.runner_from_config(config)
# out_bam, _ = broad_runner.run_fn("picard_formatconverter", out_sam)
names = {'rg': in_fastq, 'library': 'RRBS_LIB', 'pl': 'Illumina', 'pu': 'R1', 'sm': in_fastq, 'sample': sample}
out_fix_bam = broad_runner.run_fn("picard_fix_rgs", out_bam, names)
order_bam = splitext_plus(out_fix_bam)[0] + "_order.bam"
broad_runner.run_fn("picard_reorder", out_fix_bam, reference, order_bam)
index(order_bam, config)
if bowtie2:
order_bam = _set_quality(order_bam)
index(order_bam, config)
return order_bam
def _trimming(in_fastq, out_dir, sample, is_rrbs, is_directional):
"""
Trimming reads using trim_galore
"""
trim_galore = find_cmd("trim_galore")
if is_rrbs:
is_rrbs = "--rrbs"
if is_directional:
is_directional = ""
else:
is_directional = "--non_directional"
cmd = "{trim_galore} {is_rrbs} {is_directional} --length 30 --quality 30 {in_fastq} -o {tx_dir}"
trimming = op.join(out_dir, sample, sample + "_trimmed.fq")
if in_fastq.endswith("gz"):
trimming += ".gz"
with chdir(out_dir):
if not file_exists(trimming):
with tx_tmpdir() as tx_dir:
logger.debug(cmd.format(**locals()))
run(cmd.format(**locals()), "trim_galore in %s" % in_fastq)
shutil.move(tx_dir, op.join(out_dir, sample))
assert op.exists(trimming), "trimming file doesn't exists:%s" % trimming
return trimming
def _count_covars(in_bam, sample, workdir, snp, reference, config):
"""
countcovars from BisSNP tool
"""
bissnp = do.find_cmd("bissnp")
basename = sample
num_cores = config['algorithm'].get('cores', 1)
memory = config['algorithm'].get('memory', 4)
jvm_opts = "-Xms750m -Xmx%sg" % memory
cmd = ("{bissnp} {jvm_opts} -R {reference} -I {in_bam} "
"-T BisulfiteCountCovariates "
"-knownSites {snp} "
"-cov ReadGroupCovariate "
"-cov QualityScoreCovariate "
"-cov CycleCovariate "
"-recalFile {tx_out} "
"-nt {num_cores} ")
with chdir(workdir):
out_count = op.join(workdir, sample + "_recal1.csv")
if not file_exists(out_count):
with file_transaction(out_count) as tx_out:
log.logger.debug(cmd.format(**locals()))
do.run(cmd.format(**locals()), "BisSNP countcovarts in %s" % in_bam)
return out_count
