def command(self):
regions_file = "{scratch}/{uuid}.regions".format(scratch=self.scratch,
uuid=uuid.uuid4())
bed_to_regions_cmd = "cat {} | bed_to_regions.py > {}".format(
self.target_bed, regions_file)
call_somatic_cmd = " | {} -c 'from autoseq.util.bcbio import call_somatic; import sys; print call_somatic(sys.stdin.read())' ".format(
sys.executable)
freebayes_cmd = "freebayes-parallel {} {} ".format(regions_file, self.threads) + \
required("-f ", self.reference_sequence) + " --use-mapping-quality " + \
optional("--min-alternate-fraction ", self.min_alt_frac) + \
optional("--min-coverage ", self.min_coverage) + \
conditional(self.use_harmonic_indel_quals, "--harmonic-indel-quality") + \
optional("", self.params) + \
repeat(" ", self.input_bams) + \
"""| bcftools filter -i 'ALT="<*>" || QUAL > 5' """ + \
"| filter_erroneus_alt.py -V /dev/stdin " + \
conditional(self.somatic_only, call_somatic_cmd) + \
" | " + vt_split_and_leftaln(self.reference_sequence) + \
" | vcfuniq | bcftools view --apply-filters .,PASS " + \
" | bgzip > {output} && tabix -p vcf {output}".format(output=self.output)
# reason for 'vcfuniq': freebayes sometimes report duplicate variants that need to be uniqified.
rm_regions_cmd = "rm {}".format(regions_file)
return " && ".join([bed_to_regions_cmd, freebayes_cmd, rm_regions_cmd])
def command(self):
required("", self.input_tumor)
required("", self.input_normal)
freq_filter = (
" bcftools filter -e 'STATUS !~ \".*Somatic\"' 2> /dev/null "
"| %s -c 'from autoseq.util.bcbio import depth_freq_filter_input_stream; import sys; print depth_freq_filter_input_stream(sys.stdin, %s, \"%s\")' "
% (sys.executable, 0, 'bwa'))
somatic_filter = (
" sed 's/\\.*Somatic\\\"/Somatic/' " # changes \".*Somatic\" to Somatic
"| sed 's/REJECT,Description=\".*\">/REJECT,Description=\"Not Somatic via VarDict\">/' "
"| %s -c 'from autoseq.util.bcbio import call_somatic; import sys; print call_somatic(sys.stdin.read())' "
% sys.executable)
blacklist_filter = " | intersectBed -a . -b {} | ".format(
self.blacklist_bed)
cmd = "vardict-java " + required("-G ", self.reference_sequence) + \
optional("-f ", self.min_alt_frac) + \
required("-N ", self.tumorid) + \
optional("-r ", self.min_num_reads) + \
" -b \"{}|{}\" ".format(self.input_tumor, self.input_normal) + \
" -c 1 -S 2 -E 3 -g 4 -Q 10 " + required("", self.target_bed) + \
" | testsomatic.R " + \
" | var2vcf_paired.pl -P 0.9 -m 4.25 -M " + required("-f ", self.min_alt_frac) + \
" -N \"{}|{}\" ".format(self.tumorid, self.normalid) + \
" | " + freq_filter + " | " + somatic_filter + " | " + fix_ambiguous_cl() + " | " + remove_dup_cl() + \
" | vcfstreamsort -w 1000 " + \
" | " + vt_split_and_leftaln(self.reference_sequence) + \
" | bcftools view --apply-filters .,PASS " + \
" | vcfsorter.pl {} /dev/stdin ".format(self.reference_dict) + \
conditional(blacklist_filter, self.blacklist_bed) + \
" | bgzip > {output} && tabix -p vcf {output}".format(output=self.output)
return cmd
def command(self):
# activating conda env
activate_cmd = "source activate purecn-env"
# running PureCN
running_cmd = "PureCN.R " + required("--out ", self.outdir) + \
required("--sampleid ", self.tumorid) + \
required("--segfile ", self.input_seg) + \
required("--tumor ", self.input_cnr) + \
required("--vcf ", self.input_vcf) + \
required("--genome ", self.genome) + \
optional("--funsegmentation ", self.funseg) + \
optional("--minpurity ", self.minpurity) + \
optional("--hzdev ", self.hzdev) + \
optional("--maxnonclonal ", self.maxnonclonal) + \
optional("--minaf ", self.minaf) + \
optional("--error ", self.error) + \
conditional(self.postopt, "--postoptimize")
# deactivating the conda env
deactivate_cmd = "conda deactivate"
# touching required output files
touch_cmd = "touch {} {} {} {}".format(self.out_csv, self.out_genes,
self.out_variants, self.output)
return " && ".join(
[activate_cmd, running_cmd, deactivate_cmd, touch_cmd])
def command(self):
return "picard -Xmx5g -XX:ParallelGCThreads=1 " + \
required("-Djava.io.tmpdir=", self.scratch) + \
" MarkDuplicates " + \
required("INPUT=", self.input_bam) + \
required("METRICS_FILE=", self.output_metrics) + \
required("OUTPUT=", self.output_bam) + \
conditional(self.remove_duplicates, "REMOVE_DUPLICATES=true") + \
" && samtools index " + required("", self.output_bam)
def command(self):
return "gatk-klevebring -T HeterozygoteConcordance " + \
required("-R ", self.reference_sequence) + \
required("-V ", self.input_vcf) + \
required("-I ", self.input_bam) + \
required("-sid ", self.normalid) + \
optional("-L ", self.target_regions) + \
conditional(self.filter_reads_with_N_cigar, "--filter_reads_with_N_cigar") + \
required("-o ", self.output)
def command(self):
if not self.reference and not self.targets_bed:
raise ValueError("Either reference or targets_bed must be supplied")
if self.reference and self.targets_bed:
raise ValueError("Supply either reference OR targets_bed")
tmpdir = "{}/cnvkit-{}".format(self.scratch, uuid.uuid4())
sample_prefix = stripsuffix(os.path.basename(self.input_bam), ".bam")
cnvkit_cmd = "cnvkit.py batch " + required("", self.input_bam) + \
optional("-r ", self.reference) + \
conditional(self.targets_bed, "--fasta " + str(self.fasta) + " --split ") + \
conditional(self.targets_bed, "-n") + \
optional("-t ", self.targets_bed) + \
required("-d ", tmpdir)
copy_cns_cmd = "cp {}/{}.cns ".format(tmpdir, sample_prefix) + required(" ", self.output_cns)
copy_cnr_cmd = "cp {}/{}.cnr ".format(tmpdir, sample_prefix) + required(" ", self.output_cnr)
rm_cmd = "rm -r {}".format(tmpdir)
return " && ".join([cnvkit_cmd, copy_cns_cmd, copy_cnr_cmd, rm_cmd])
def command(self):
tmpdir = "{}/write-alascca-report-{}".format(self.scratch,
uuid.uuid4())
mkdir_tmp_cmd = "mkdir -p {}".format(tmpdir)
tmp_pdf = os.path.join(tmpdir, 'Report.pdf')
cmd = 'writeAlasccaReport ' + \
required(' --tmp_dir ', tmpdir) + \
required(' --output_dir ', tmpdir) + \
conditional(self.alascca_only, " --alascca_only ") + \
required('', self.input_genomic_json) + \
required('', self.input_metadata_json)
cp_cmd = "cp {} {}".format(tmp_pdf, self.output_pdf)
rmdir_cmd = "rm -r {}".format(tmpdir)
return " && ".join([mkdir_tmp_cmd, cmd, cp_cmd, rmdir_cmd])
def command(self):
filt_vcf = "{scratch}/{uuid}.vcf.gz".format(scratch=self.scratch,
uuid=uuid.uuid4())
bgzip = ""
tabix = ""
if self.output.endswith('gz'):
bgzip = "| bgzip"
tabix = " && tabix -p vcf {}".format(self.output)
filt_vcf_cmd = "vcf_filter.py --no-filtered " + required("", self.input_vcf) + " sq --site-quality 5 " + \
"|bgzip" + " > " + filt_vcf
vcf_add_sample_cmd = "vcf_add_sample.py " + \
conditional(self.filter_hom, "--filter_hom") + \
required("--samplename ", self.samplename) + \
filt_vcf + " " + \
required("", self.input_bam) + \
bgzip + " > " + self.output + tabix
rm_filt_cmd = "rm " + filt_vcf
return " && ".join([filt_vcf_cmd, vcf_add_sample_cmd, rm_filt_cmd])
def command(self):
# activating conda env
activate_cmd = "source activate purecn-env"
# running PureCN
running_cmd = "PureCN.R " + required("--out ", self.outdir) + \
required("--sampleid ", self.tumorid) + \
required("--segfile ", self.input_seg) + \
required("--vcf ", self.input_vcf) + \
required("--gcgene ", self.gcgene_file) + \
required("--genome ", self.genome) + \
optional("--minpurity ", self.minpurity) + \
optional("--hzdev ", self.hzdev) + \
optional("--segfilesdev ", self.seg_sdev) + \
conditional(self.postopt, "--postoptimize")
# deactivating the conda env
deactivate_cmd = "source deactivate"
return " && ".join([activate_cmd, running_cmd, deactivate_cmd])
def command(self):
return "fastqc -o {} ".format(self.outdir) + \
conditional(self.extract, "--extract") + \
" --nogroup {}".format(self.input)
def vt_split_and_leftaln(reference_sequence, allow_ref_mismatches=False):
cmd = "vt decompose -s - " + \
"|vt normalize " + conditional(allow_ref_mismatches, ' -n ') + \
" -r {} - ".format(reference_sequence)
return cmd
def command(self):
return "msisensor scan " + \
required("-d ", self.input_fasta) + \
required("-o ", self.output) + \
conditional(self.homopolymers_only, " -p 1 ")