def add_gatk_variantcaller(self, normal_bam_source, tumor_bam_source):
if "generate_gatk_intervals" in self.step_nodes:
generated_intervals = self.generate_gatk_intervals.out_regions
else:
generated_intervals = self.step(
"generate_gatk_intervals",
GenerateIntervalsByChromosome(reference=self.reference),
when=self.gatk_intervals.is_null(),
).out_regions
intervals = FirstOperator([self.gatk_intervals, generated_intervals])
recal_ins = {
"reference": self.reference,
"intervals": intervals,
"snps_dbsnp": self.snps_dbsnp,
"snps_1000gp": self.snps_1000gp,
"known_indels": self.known_indels,
"mills_indels": self.mills_indels,
}
self.step(
"bqsr_normal",
GATKBaseRecalBQSRWorkflow_4_1_3(bam=normal_bam_source,
**recal_ins),
scatter="intervals",
)
self.step(
"bqsr_tumor",
GATKBaseRecalBQSRWorkflow_4_1_3(bam=tumor_bam_source, **recal_ins),
scatter="intervals",
)
self.step(
"vc_gatk",
GatkSomaticVariantCaller_4_1_3(
normal_bam=self.bqsr_normal.out,
tumor_bam=self.bqsr_tumor.out,
normal_name=self.normal_name,
intervals=intervals,
reference=self.reference,
gnomad=self.gnomad,
panel_of_normals=self.panel_of_normals,
),
scatter=["intervals", "normal_bam", "tumor_bam"],
)
self.step("vc_gatk_merge",
Gatk4GatherVcfs_4_1_3(vcfs=self.vc_gatk.out))
self.step("vc_gatk_compressvcf",
BGZipLatest(file=self.vc_gatk_merge.out))
self.step(
"vc_gatk_sort_combined",
BcfToolsSort_1_9(
vcf=self.vc_gatk_compressvcf.out.as_type(CompressedVcf)),
)
self.step(
"vc_gatk_uncompressvcf",
UncompressArchive(file=self.vc_gatk_sort_combined.out),
)
def add_vardict_variantcaller(self, normal_bam_source, tumor_bam_source):
self.step(
"generate_vardict_headerlines",
GenerateVardictHeaderLines(reference=self.reference),
)
self.step(
"vc_vardict",
VardictSomaticVariantCaller(
normal_bam=normal_bam_source,
tumor_bam=tumor_bam_source,
normal_name=self.normal_name,
tumor_name=self.tumor_name,
header_lines=self.generate_vardict_headerlines.out,
intervals=self.vardict_intervals,
reference=self.reference,
allele_freq_threshold=self.allele_freq_threshold,
minMappingQual=self.minMappingQual,
filter=self.filter,
),
scatter="intervals",
)
self.step("vc_vardict_merge",
Gatk4GatherVcfs_4_1_3(vcfs=self.vc_vardict.out))
self.step("vc_vardict_compress_for_sort",
BGZipLatest(file=self.vc_vardict_merge.out))
self.step(
"vc_vardict_sort_combined",
BcfToolsSort_1_9(vcf=self.vc_vardict_compress_for_sort.out.as_type(
CompressedVcf)),
)
self.step(
"vc_vardict_uncompress_for_combine",
UncompressArchive(file=self.vc_vardict_sort_combined.out),
)
self.output(
"out_variants_vardict",
source=self.vc_vardict_sort_combined.out,
output_folder=[
"vcf",
],
output_name=StringFormatter(
"{tumor_name}--{normal_name}_vardict",
tumor_name=self.tumor_name,
normal_name=self.normal_name,
),
doc="Merged variants from the VarDict caller",
)
self.output(
"out_variants_vardict_split",
source=self.vc_vardict.out,
output_folder=[
"vcf",
"VardictByInterval",
],
doc="Unmerged variants from the GATK caller (by interval)",
)
def add_vardict_variantcaller(self, bam_source):
self.input(
"allele_freq_threshold",
Float,
0.05,
),
self.input("minMappingQual", Int(optional=True))
self.input("filter", String(optional=True))
# Vardict
self.step(
"generate_vardict_headerlines",
GenerateVardictHeaderLines(reference=self.reference),
)
self.step(
"vc_vardict",
VardictGermlineVariantCaller(
bam=bam_source,
reference=self.reference,
intervals=self.vardict_intervals,
sample_name=self.sample_name,
allele_freq_threshold=self.allele_freq_threshold,
header_lines=self.generate_vardict_headerlines.out,
minMappingQual=self.minMappingQual,
filter=self.filter,
),
scatter="intervals",
)
self.step("vc_vardict_merge", Gatk4GatherVcfs_4_1_3(vcfs=self.vc_vardict.out))
self.step(
"vc_vardict_compress_for_sort",
BGZipLatest(file=self.vc_vardict_merge.out.as_type(Vcf)),
)
self.step(
"vc_vardict_sort_combined",
BcfToolsSort_1_9(
vcf=self.vc_vardict_compress_for_sort.out.as_type(CompressedVcf)
),
)
self.step(
"vc_vardict_uncompress_for_combine",
UncompressArchive(file=self.vc_vardict_sort_combined.out),
)
self.output(
"out_variants_vardict",
source=self.vc_vardict_sort_combined.out,
output_folder=["variants"],
output_name="vardict",
doc="Merged variants from the VarDict caller",
)
self.output(
"out_variants_vardict_split",
source=self.vc_vardict.out,
output_folder=["variants", "vardict"],
doc="Unmerged variants from the VarDict caller (by interval)",
)
def add_combine_variants(self, normal_bam_source, tumor_bam_source):
self.step(
"combine_variants",
CombineVariants_0_0_8(
normal=self.normal_name,
tumor=self.tumor_name,
vcfs=[
self.vc_gatk_uncompressvcf.out.as_type(Vcf),
self.vc_strelka.out,
self.vc_vardict_uncompress_for_combine.out.as_type(Vcf),
],
type="somatic",
columns=["AD", "DP", "GT"],
),
)
self.step("combined_compress",
BGZipLatest(file=self.combine_variants.out))
self.step(
"combined_sort",
BcfToolsSort_1_9(
vcf=self.combined_compress.out.as_type(CompressedVcf)),
)
self.step("combined_uncompress",
UncompressArchive(file=self.combined_sort.out))
self.step(
"combined_addbamstats",
AddBamStatsSomatic_0_1_0(
normal_id=self.normal_name,
tumor_id=self.tumor_name,
normal_bam=normal_bam_source,
tumor_bam=tumor_bam_source,
vcf=self.combined_uncompress.out.as_type(Vcf),
reference=self.reference,
),
)
self.output(
"out_variants",
source=self.combined_addbamstats.out,
output_folder=[
"vcf",
],
output_name=StringFormatter(
"{tumor_name}--{normal_name}_combined",
tumor_name=self.tumor_name,
normal_name=self.normal_name,
),
doc="Combined variants from GATK, VarDict and Strelka callers",
)
def constructor(self):
self.input("bam", BamBai)
self.input("reference", FastaWithDict)
self.input("intervals", BedTabix(optional=True))
self.input("is_exome", Boolean(optional=True))
self.step(
"manta",
Manta_1_5_0(
bam=self.bam,
reference=self.reference,
callRegions=self.intervals,
exome=self.is_exome,
),
)
self.step(
"strelka",
StrelkaGermline_2_9_10(
bam=self.bam,
reference=self.reference,
indelCandidates=self.manta.candidateSmallIndels,
callRegions=self.intervals,
exome=self.is_exome,
),
)
# normalise and filter "PASS" variants
self.step("uncompressvcf",
UncompressArchive(file=self.strelka.variants))
self.step(
"splitnormalisevcf",
SplitMultiAllele(vcf=self.uncompressvcf.out,
reference=self.reference),
)
self.step(
"filterpass",
VcfToolsvcftoolsLatest(
vcf=self.splitnormalisevcf.out,
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
self.output("sv", source=self.manta.diploidSV)
self.output("variants", source=self.strelka.variants)
self.output("out", source=self.filterpass.out)
def add_combine_variants(self, bam_source):
# Note, this is reliant on the specific step names from previous steps
# Combine
self.step(
"combine_variants",
CombineVariants_0_0_8(
vcfs=[
self.vc_gatk_uncompress.out.as_type(Vcf),
self.vc_strelka.out,
self.vc_vardict_uncompress_for_combine.out.as_type(Vcf),
],
type="germline",
columns=["AC", "AN", "AF", "AD", "DP", "GT"],
),
)
self.step("combined_compress",
BGZipLatest(file=self.combine_variants.out))
self.step(
"combined_sort",
BcfToolsSort_1_9(
vcf=self.combined_compress.out.as_type(CompressedVcf)),
)
self.step("combined_uncompress",
UncompressArchive(file=self.combined_sort.out))
self.step(
"combined_addbamstats",
AddBamStatsGermline_0_1_0(
bam=bam_source,
vcf=self.combined_uncompress.out.as_type(Vcf),
reference=self.reference,
),
)
self.output(
"out_variants",
source=self.combined_addbamstats.out,
output_folder="variants",
output_name="combined",
doc="Combined variants from all 3 callers",
)
def constructor(self):
self.input("bam", BamBai)
self.input(
"intervals",
Bed(optional=True),
doc=
"This optional interval supports processing by regions. If this input resolves "
"to null, then GATK will process the whole genome per each tool's spec",
)
self.input("reference", FastaWithDict)
self.input("snps_dbsnp", VcfTabix)
self.step(
"split_bam",
gatk4.Gatk4SplitReads_4_1_3(bam=self.bam,
intervals=self.intervals),
)
self.step(
"haplotype_caller",
gatk4.Gatk4HaplotypeCaller_4_1_3(
inputRead=self.split_bam.out,
intervals=self.intervals,
reference=self.reference,
dbsnp=self.snps_dbsnp,
pairHmmImplementation="LOGLESS_CACHING",
),
)
self.step("uncompressvcf",
UncompressArchive(file=self.haplotype_caller.out))
self.step(
"splitnormalisevcf",
SplitMultiAllele(vcf=self.uncompressvcf.out,
reference=self.reference),
)
self.output("variants", source=self.haplotype_caller.out)
self.output("out_bam", source=self.haplotype_caller.bam)
self.output("out", source=self.splitnormalisevcf.out)
def constructor(self):
self.input("normal_bam", BamBai)
self.input("tumor_bam", BamBai)
self.input("normal_name", String(optional=True))
self.input(
"intervals",
Bed(optional=True),
doc=
"This optional intervals file supports processing by regions. If this file resolves "
"to null, then GATK will process the whole genome per each tool's spec",
)
self.input("reference", FastaWithDict)
self.input("gnomad", VcfTabix)
self.input("panel_of_normals", VcfTabix(optional=True))
# split normal and tumor bam
self.step(
"normal_split_bam",
self.process_subpipeline(bam=self.normal_bam,
intervals=self.intervals),
)
self.step(
"tumor_split_bam",
self.process_subpipeline(bam=self.tumor_bam,
intervals=self.intervals),
)
# variant calling + learn read orientation model
self.step(
"mutect2",
gatk4.GatkMutect2_4_1_3(
normalBams=[self.normal_split_bam.out],
tumorBams=[self.tumor_split_bam.out],
normalSample=self.normal_name,
intervals=self.intervals,
reference=self.reference,
germlineResource=self.gnomad,
panelOfNormals=self.panel_of_normals,
),
)
self.step(
"learnorientationmodel",
gatk4.Gatk4LearnReadOrientationModelLatest(
f1r2CountsFiles=self.mutect2.f1f2r_out, ),
)
# calculate contamination and segmentation
self.step(
"getpileupsummaries",
gatk4.Gatk4GetPileUpSummariesLatest(
bam=self.tumor_split_bam.out,
sites=self.gnomad,
intervals=self.intervals,
),
)
self.step(
"calculatecontamination",
gatk4.Gatk4CalculateContaminationLatest(
pileupTable=self.getpileupsummaries.out, ),
)
self.step(
"filtermutect2calls",
gatk4.Gatk4FilterMutectCallsLatest(
vcf=self.mutect2.out,
reference=self.reference,
segmentationFile=self.calculatecontamination.segOut,
contaminationTable=self.calculatecontamination.contOut,
readOrientationModel=self.learnorientationmodel.out,
statsFile=self.mutect2.stats,
),
)
# normalise and filter "PASS" variants
self.step("uncompressvcf",
UncompressArchive(file=self.filtermutect2calls.out))
self.step(
"splitnormalisevcf",
SplitMultiAllele(vcf=self.uncompressvcf.out,
reference=self.reference),
)
self.step(
"filterpass",
VcfToolsvcftoolsLatest(
vcf=self.splitnormalisevcf.out,
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
self.output("variants", source=self.filtermutect2calls.out)
self.output("out_bam", source=self.mutect2.bam)
self.output("out", source=self.filterpass.out)
def add_gatk_variantcaller(self, bam_source):
# VARIANT CALLERS
intervals = FirstOperator(
[
self.gatk_intervals,
self.step(
"generate_gatk_intervals",
GenerateIntervalsByChromosome(reference=self.reference),
when=self.gatk_intervals.is_null(),
).out_regions,
]
)
# GATK
self.step(
"bqsr",
GATKBaseRecalBQSRWorkflow_4_1_3(
bam=bam_source,
reference=self.reference,
snps_dbsnp=self.snps_dbsnp,
snps_1000gp=self.snps_1000gp,
known_indels=self.known_indels,
mills_indels=self.mills_indels,
intervals=intervals,
),
scatter="intervals",
doc="Perform base quality score recalibration",
)
self.step(
"vc_gatk",
GatkGermlineVariantCaller_4_1_3(
bam=self.bqsr.out,
intervals=intervals,
reference=self.reference,
snps_dbsnp=self.snps_dbsnp,
),
scatter=["intervals", "bam"],
)
self.step("vc_gatk_merge", Gatk4GatherVcfs_4_1_3(vcfs=self.vc_gatk.out))
self.step("vc_gatk_compressvcf", BGZipLatest(file=self.vc_gatk_merge.out))
self.step(
"vc_gatk_sort_combined",
BcfToolsSort_1_9(vcf=self.vc_gatk_compressvcf.out.as_type(CompressedVcf)),
)
self.step(
"vc_gatk_uncompress",
UncompressArchive(file=self.vc_gatk_sort_combined.out),
)
self.output(
"out_variants_gatk",
source=self.vc_gatk_sort_combined.out,
output_folder="variants",
output_name="gatk",
doc="Merged variants from the GATK caller",
)
self.output(
"out_variants_gatk_split",
source=self.vc_gatk.out,
output_folder=["variants", "gatk"],
doc="Unmerged variants from the GATK caller (by interval)",
)
def add_gatk_variantcaller(self, normal_bam_source, tumor_bam_source):
"""
Reimplemented because need steps for combine
"""
if "generate_gatk_intervals" in self.step_nodes:
generated_intervals = self.generate_gatk_intervals.out_regions
else:
generated_intervals = self.step(
"generate_gatk_intervals",
GenerateIntervalsByChromosome(reference=self.reference),
when=self.gatk_intervals.is_null(),
).out_regions
intervals = FirstOperator([self.gatk_intervals, generated_intervals])
recal_ins = {
"reference": self.reference,
"intervals": intervals,
"snps_dbsnp": self.snps_dbsnp,
"snps_1000gp": self.snps_1000gp,
"known_indels": self.known_indels,
"mills_indels": self.mills_indels,
}
self.step(
"bqsr_normal",
GATKBaseRecalBQSRWorkflow_4_1_3(bam=normal_bam_source,
**recal_ins),
scatter="intervals",
)
self.step(
"bqsr_tumor",
GATKBaseRecalBQSRWorkflow_4_1_3(bam=tumor_bam_source, **recal_ins),
scatter="intervals",
)
self.step(
"vc_gatk",
GatkSomaticVariantCaller_4_1_3(
normal_bam=self.bqsr_normal.out,
tumor_bam=self.bqsr_tumor.out,
normal_name=self.normal_name,
intervals=intervals,
reference=self.reference,
gnomad=self.gnomad,
panel_of_normals=self.panel_of_normals,
),
scatter=["intervals", "normal_bam", "tumor_bam"],
)
self.step("vc_gatk_merge",
Gatk4GatherVcfs_4_1_3(vcfs=self.vc_gatk.out))
self.step("vc_gatk_compress_for_sort",
BGZipLatest(file=self.vc_gatk_merge.out))
self.step(
"vc_gatk_sort_combined",
BcfToolsSort_1_9(
vcf=self.vc_gatk_compress_for_sort.out.as_type(CompressedVcf)),
)
self.step(
"vc_gatk_uncompressvcf",
UncompressArchive(file=self.vc_gatk_sort_combined.out),
)
# VCF
self.output(
"out_variants_gatk",
source=self.vc_gatk_sort_combined.out,
output_folder=[
"vcf",
],
output_name=StringFormatter(
"{tumor_name}--{normal_name}_gatk",
tumor_name=self.tumor_name,
normal_name=self.normal_name,
),
doc="Merged variants from the GATK caller",
)
self.output(
"out_variants_split",
source=self.vc_gatk.out,
output_folder=[
"vcf",
"GATKByInterval",
],
doc="Unmerged variants from the GATK caller (by interval)",
)
def constructor(self):
##INPUTS
self.input("bam", BamBai())
self.input("sample_name", String())
self.input("reference_folder", Directory())
self.input("intervals", Bed())
self.input("gemini_chromosomes", String(optional=True))
self.input("ploidy", String(optional=True), default="somatic")
self.input("min_bq", Int(optional=True))
self.input("min_mq", Int(optional=True))
self.input("min_dp", Int(optional=True))
self.input("min_vaf", Float(optional=True))
self.input("vc_min_vq", Int(optional=True))
self.input("noise_level", Int(optional=True))
self.input("vqr_min_vq", Int(optional=True))
self.input("pisces_awk_script", File())
## STEPS
self.step(
"primary_only",
SamToolsView_1_9(sam=self.bam,
doNotOutputAlignmentsWithBitsSet="0x100"),
)
self.step(
"index_primary_only_bam",
SamToolsIndex_1_9(bam=self.primary_only.out),
)
self.step(
"gemini_read_preprocessing",
PiscesGemini_5_3_0_0(
inputBam=self.index_primary_only_bam,
referenceFolder=self.reference_folder,
samtoolsExecutable="samtools",
chromosomeFilter=self.gemini_chromosomes,
outputDir=".",
piscesVersion="5.3.0.0",
),
)
self.step(
"pisces",
PiscesVariantCaller_5_3_0_0(
inputBam=self.gemini_read_preprocessing.bam,
referenceFolder=self.reference_folder,
outputDir=".",
intervalBedFile=self.intervals,
ploidy=self.ploidy,
minimumBaseQuality=self.min_bq,
minimumMappingQuality=self.min_mq,
minimumVariantFrequency=self.min_vaf,
noiseLevelForQModel=self.noise_level,
minimumVariantFrequencyFilter=self.min_vaf,
enableSingleStrandFilter="True",
outputSBFiles="True",
callMNVs="False",
maxMNVLength=1,
RMxNFilter="5,9,0.35",
variantQualityFilter=self.vc_min_vq,
crushVCF="False",
gVCF="False",
piscesVersion="5.3.0.0",
),
)
self.step(
"vqr",
PiscesVariantQualityRecalibration_5_3_0_0(
inputVcf=self.pisces.vcf,
outputDir=".",
baselineNoise=self.noise_level,
minVariantQuality=self.vqr_min_vq,
piscesVersion="5.3.0.0",
),
)
piscesVcf = FirstOperator([self.vqr.vcf, self.pisces.vcf])
self.step(
"fixSource",
Awk(script=self.pisces_awk_script, input_files=piscesVcf),
)
self.step("sort", BcfToolsSort_1_9(vcf=self.fixSource.out))
self.step("normalise", BcfToolsNorm_1_9(vcf=self.sort.out))
self.step("uncompress", UncompressArchive(file=self.normalise.out))
self.step(
"filterpass",
VcfToolsvcftools_0_1_16(
vcf=self.uncompress.out.as_type(Vcf),
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
## OUTPUTS
self.output("variants", source=self.sort.out)
self.output("out", source=self.filterpass.out)
self.output("out_bam", source=self.gemini_read_preprocessing.bam)
def constructor(self):
# Inputs
self.input("sample_name", String)
self.input("fastqs", Array(FastqGzPair))
self.input("seqrun", String, doc="SeqRun Name (for Vcf2Tsv)")
self.input("reference", FastaWithDict)
self.input("region_bed", Bed)
self.input("region_bed_extended", Bed)
self.input("region_bed_annotated", Bed)
self.input("genecoverage_bed", Bed)
self.input("genome_file", TextFile)
self.input("panel_name", String)
self.input("vcfcols", TextFile)
self.input("black_list", Bed(optional=True))
self.input("snps_dbsnp", VcfTabix)
self.input("snps_1000gp", VcfTabix)
self.input("known_indels", VcfTabix)
self.input("mills_indels", VcfTabix)
self.input("mutalyzer_server", String)
self.input("pathos_db", String)
self.input("maxRecordsInRam", Int)
# tumor only
self.input("gnomad", VcfTabix)
self.input("panel_of_normals", VcfTabix(optional=True))
# fastqc
self.step(
"fastqc", FastQC_0_11_5(reads=self.fastqs, threads=4), scatter="reads"
)
# get the overrepresentative sequence from fastqc
self.step(
"getfastqc_adapters",
ParseFastqcAdaptors(fastqc_datafiles=self.fastqc.datafile,),
scatter="fastqc_datafiles",
)
# align and generate sorted index bam
self.step(
"align_and_sort",
BwaAligner(
fastq=self.fastqs,
reference=self.reference,
sample_name=self.sample_name,
sortsam_tmpDir=".",
cutadapt_adapter=self.getfastqc_adapters,
cutadapt_removeMiddle3Adapter=self.getfastqc_adapters,
),
scatter=["fastq", "cutadapt_adapter", "cutadapt_removeMiddle3Adapter"],
)
# merge into one bam and markdups
self.step(
"merge_and_mark",
MergeAndMarkBams_4_1_3(
bams=self.align_and_sort.out,
sampleName=self.sample_name,
maxRecordsInRam=self.maxRecordsInRam,
),
)
# performance: doc
self.step(
"annotate_doc",
AnnotateDepthOfCoverage_0_1_0(
bam=self.merge_and_mark.out,
bed=self.region_bed_annotated,
reference=self.reference,
sample_name=self.sample_name,
),
)
# performance
self.step(
"performance_summary",
PerformanceSummaryTargeted_0_1_0(
bam=self.merge_and_mark.out,
region_bed=self.region_bed,
genecoverage_bed=self.genecoverage_bed,
sample_name=self.sample_name,
genome_file=self.genome_file,
),
)
# gridss
self.step(
"gridss",
Gridss_2_6_2(
bams=self.merge_and_mark.out,
reference=self.reference,
blacklist=self.black_list,
tmpdir=".",
),
)
# post gridss r for tumor only + tumor only mode
# self.step("gridss_post_r", GRIDSSProcessOutput(inp=self.gridss.out))
# gatk bqsr bam
self.step(
"bqsr",
GATKBaseRecalBQSRWorkflow_4_1_3(
bam=self.merge_and_mark.out,
intervals=self.region_bed_extended,
reference=self.reference,
snps_dbsnp=self.snps_dbsnp,
snps_1000gp=self.snps_1000gp,
known_indels=self.known_indels,
mills_indels=self.mills_indels,
),
)
# mutect2
self.step(
"mutect2",
GatkSomaticVariantCallerTumorOnlyTargeted(
bam=self.bqsr.out,
intervals=self.region_bed_extended,
reference=self.reference,
gnomad=self.gnomad,
panel_of_normals=self.panel_of_normals,
),
)
# haplotypecaller to do: take base recal away from the
self.step(
"haplotype_caller",
Gatk4HaplotypeCaller_4_1_3(
inputRead=self.bqsr.out,
intervals=self.region_bed_extended,
reference=self.reference,
dbsnp=self.snps_dbsnp,
pairHmmImplementation="LOGLESS_CACHING",
),
)
self.step(
"splitnormalisevcf",
SplitMultiAlleleNormaliseVcf(
compressedVcf=self.haplotype_caller.out, reference=self.reference
),
)
# combine variants
self.step(
"combinevariants",
CombineVariants_0_0_8(
vcfs=[self.splitnormalisevcf.out, self.mutect2.out],
type="germline",
columns=["AD", "DP", "AF", "GT"],
),
)
self.step("compressvcf", BGZip_1_9(file=self.combinevariants.out))
self.step("sortvcf", BcfToolsSort_1_9(vcf=self.compressvcf.out))
self.step("uncompressvcf", UncompressArchive(file=self.sortvcf.out))
# addbamstats
self.step(
"addbamstats",
AddBamStatsGermline_0_1_0(
bam=self.merge_and_mark.out,
vcf=self.uncompressvcf.out,
reference=self.reference,
),
)
# Molpath specific processes
self.step("compressvcf2", BGZip_1_9(file=self.addbamstats.out))
self.step("tabixvcf", TabixLatest(inp=self.compressvcf2.out))
self.step(
"calculate_variant_length",
VcfLength_1_0_1(vcf=self.tabixvcf.out),
doc="Add the length column for the output of AddBamStats",
)
filter_for_variants = self.input("filter_for_vcfs", str, default="length > 150")
self.step(
"filter_variants_1_failed",
VcfFilter_1_0_1(
vcf=self.calculate_variant_length.out, info_filter=filter_for_variants
),
)
self.step(
"filter_variants_1",
VcfFilter_1_0_1(
vcf=self.calculate_variant_length.out,
info_filter=filter_for_variants,
invert=True, # -v param
),
)
# Jiaan: copy over from the FRCP, can take the block comment out
# # This one is the in-house molpath step
# self.step(
# "normalise_vcfs",
# NormaliseVcf_1_5_4(
# pathos_version=self.pathos_db,
# mutalyzer=self.mutalyzer_server, # mutalyzer="https://vmpr-res-mutalyzer1.unix.petermac.org.au",
# rdb=self.pathos_db, # rdb="pa_uat",
# inp=self.filter_variants_1.out,
# ),
# )
# # repeat remove 150bp variants (workaround for normalise_vcf bug)
# self.step(
# "filter_variants_2_failed",
# VcfFilter_1_0_1(
# vcf=self.normalise_vcfs.out, info_filter=filter_for_variants
# ),
# )
# self.step(
# "filter_variants_2",
# VcfFilter_1_0_1(
# vcf=self.normalise_vcfs.out,
# info_filter=filter_for_variants,
# invert=True, # -v param
# ),
# )
# self.step(
# "convert_to_tsv",
# Vcf2Tsv_1_5_4(
# pathos_version=self.pathos_db,
# inp=self.filter_variants_2.out,
# sample=self.sample_name,
# columns=self.vcfcols,
# seqrun=self.seqrun,
# ),
# )
# self.step(
# "index_with_igvtools", IgvIndexFeature_2_5_3(inp=self.filter_variants_2.out)
# )
# output
self.output("fastq_qc", source=self.fastqc.out, output_folder="QC")
self.output("markdups_bam", source=self.merge_and_mark.out, output_folder="BAM")
self.output(
"doc_out", source=self.annotate_doc.out, output_folder="PERFORMANCE"
)
self.output(
"summary", source=self.performance_summary.out, output_folder="PERFORMANCE"
)
self.output(
"gene_summary",
source=self.performance_summary.geneFileOut,
output_folder="PERFORMANCE",
)
self.output(
"region_summary",
source=self.performance_summary.regionFileOut,
output_folder="PERFORMANCE",
)
self.output("gridss_vcf", source=self.gridss.out, output_folder="SV")
self.output("gridss_bam", source=self.gridss.assembly, output_folder="SV")
self.output(
"haplotypecaller_vcf",
source=self.haplotype_caller.out,
output_folder="VCF",
)
self.output(
"haplotypecaller_bam",
source=self.haplotype_caller.bam,
output_folder="VCF",
)
self.output(
"haplotypecaller_norm",
source=self.splitnormalisevcf.out,
output_folder="VCF",
)
self.output("mutect2_vcf", source=self.mutect2.variants, output_folder="VCF")
self.output("mutect2_bam", source=self.mutect2.out_bam, output_folder="VCF")
self.output("mutect2_norm", source=self.mutect2.out, output_folder="VCF")
self.output("addbamstats_vcf", source=self.addbamstats.out)
def constructor(self):
## INPUTS
self.input("bam", BamBai())
self.input("sample_name", String())
self.input("reference_folder", Directory())
self.input("intervals", Bed())
self.input("ploidy", String(optional=True), default="somatic")
self.input("min_bq", Int(optional=True))
self.input("min_mq", Int(optional=True))
self.input("min_dp", Int(optional=True), default=100)
self.input("min_vaf", Float(optional=True))
self.input("vc_min_vq", Int(optional=True))
self.input("noise_level", Int(optional=True))
self.input("vqr_min_vq", Int(optional=True))
self.input("pisces_awk_script", File())
## STEPS
self.step(
"primary_only",
SamToolsView_1_9(sam=self.bam,
doNotOutputAlignmentsWithBitsSet="0x100"),
)
self.step(
"index_primary_only_bam",
SamToolsIndex_1_9(bam=self.primary_only.out),
)
self.step(
"hygea_realignment",
PiscesHygeaRealigner_5_2_10_49(
inputBam=self.index_primary_only_bam,
outputDir=".",
referenceFolder=self.reference_folder,
skipAndRemoveDuplicates="true",
piscesVersion="5.2.10.49",
),
)
self.step(
"stitcher_read_joining",
PiscesStitcher_5_2_10_49(
inputBam=self.hygea_realignment.out,
outputDir=".",
sampleName=self.sample_name,
piscesVersion="5.2.10.49",
),
)
self.step(
"stitcher_sort",
SamToolsSort_1_9(
bam=self.stitcher_read_joining.out,
outputFilename=self.sample_name + ".bam",
),
)
self.step("stitcher_index",
SamToolsIndex_1_9(bam=self.stitcher_sort.out))
self.step(
"pisces",
PiscesVariantCaller_5_2_10_49(
inputBam=self.stitcher_index.out,
referenceFolder=self.reference_folder,
outputDir=".",
intervalBedFile=self.intervals,
ploidy=self.ploidy,
minimumBaseQuality=self.min_bq,
minimumMappingQuality=self.min_mq,
minimumVariantFrequency=self.min_vaf,
minimumCoverage=self.min_dp,
noiseLevelForQModel=self.noise_level,
minimumVariantFrequencyFilter=self.min_vaf,
enableSingleStrandFilter="true",
callMNVs="false",
maxMNVLength=1,
RMxNFilter="5,9,0.35",
variantQualityFilter=self.vc_min_vq,
crushVCF="false",
gVCF="false",
piscesVersion="5.2.10.49",
),
)
self.step(
"vqr",
PiscesVariantQualityRecalibration_5_2_10_49(
inputVcf=self.pisces.vcf,
outputDir=".",
baselineNoise=self.noise_level,
minVariantQuality=self.vqr_min_vq,
piscesVersion="5.2.10.49",
),
)
piscesVcf = FirstOperator([self.vqr.vcf, self.pisces.vcf])
self.step(
"fixSource",
Awk(script=self.pisces_awk_script, input_files=piscesVcf),
)
self.step("sort", BcfToolsSort_1_9(vcf=self.fixSource.out))
self.step("normalise", BcfToolsNorm_1_9(vcf=self.sort.out))
self.step("uncompress", UncompressArchive(file=self.normalise.out))
self.step(
"filterpass",
VcfToolsvcftools_0_1_16(
vcf=self.uncompress.out.as_type(Vcf),
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
## OUTPUTs
self.output("variants", source=self.sort.out)
self.output("out", source=self.filterpass.out)
self.output("out_bam", source=self.stitcher_index.out)
## OPTIONAL OUTPUTs
self.output("hygea_options",
source=self.hygea_realignment.used_options)
self.output("stitcher_options",
source=self.stitcher_read_joining.used_options)
self.output("pisces_options", source=self.pisces.used_options)
self.output("vqr_options", source=self.vqr.used_options)
def constructor(self):
## INPUTs
self.input("sample_name", String())
self.input("reference", Fasta())
# For mpileup
self.input("bam", BamBai())
self.input("pileup_max_depth", Int(optional=True))
self.input("min_bq", Int(optional=True), default=20)
# For Varscan2
self.input("min_dp", Int(optional=True))
self.input("min_ad", Int(optional=True))
self.input("min_vaf", Float(optional=True))
self.input("pval", Float(optional=True), default=0.0001)
# For correction of VCF header
self.input("header_lines", File)
## STEPS
self.step(
"mpileup",
SamToolsMpileup_1_9(
bam=self.bam,
outputFilename="./" + self.sample_name + ".mpileup",
noBAQ=True,
maxDepth=self.pileup_max_depth,
reference=self.reference,
),
)
self.step(
"varscan2cns",
VarscanMpileup2cns_2_4_2(
mpileup=self.mpileup.out,
minCoverage=self.min_dp,
minVariantReads=self.min_ad,
minBaseQuality=self.min_bq,
minVariantFrequency=self.min_vaf,
pValue=self.pval,
outputVcfFormat=1,
variantPositionsOnly=1,
),
)
self.step(
"VSheader",
VarscanHeader(
inputVcf=self.varscan2cns.out,
outputVcf="./" + self.sample_name + "VS.reheader.vcf",
),
)
self.step(
"VSheaderContigs",
BcfToolsAnnotate_1_5(vcf=self.VSheader.out,
headerLines=self.header_lines),
)
self.step("sortVcf", BcfToolsSort_1_9(vcf=self.VSheaderContigs.out))
self.step("normaliseVcf", BcfToolsNorm_1_9(vcf=self.sortVcf.out))
self.step("uncompress", UncompressArchive(file=self.normaliseVcf.out))
self.step(
"filterpass",
VcfToolsvcftools_0_1_16(
vcf=self.uncompress.out.as_type(Vcf),
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
## OUTPUTS
self.output("variants", source=self.sortVcf.out)
self.output("out", source=self.filterpass.out)