def constructor(self):
self.input("normal_bam", BamBai)
self.input("tumor_bam", BamBai)
self.input("normal_name", str)
self.input("tumor_name", str)
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("snps_dbsnp", VcfTabix)
self.input("snps_1000gp", VcfTabix)
self.input("known_indels", VcfTabix)
self.input("mills_indels", VcfTabix)
self.step(
"base_recalibrator_normal",
gatk4.Gatk4BaseRecalibrator_4_1_3(),
ignore_missing=True,
)
self.step(
"base_recalibrator_tumor",
gatk4.Gatk4BaseRecalibrator_4_1_3(),
ignore_missing=True,
)
self.step(
"apply_bqsr_normal", gatk4.Gatk4ApplyBqsr_4_1_3(), ignore_missing=True
)
self.step("apply_bqsr_tumor", gatk4.Gatk4ApplyBqsr_4_1_3(), ignore_missing=True)
# S1: BaseRecalibrator(s)
for inp, base_recal, applyBQSR in [
(self.normal_bam, self.base_recalibrator_normal, self.apply_bqsr_normal),
(self.tumor_bam, self.base_recalibrator_tumor, self.apply_bqsr_tumor),
]:
base_recal["bam"] = inp
base_recal["intervals"] = self.intervals
base_recal["reference"] = self.reference
base_recal["knownSites"] = [
self.snps_dbsnp,
self.snps_1000gp,
self.known_indels,
self.mills_indels,
]
applyBQSR["recalFile"] = base_recal.out
applyBQSR["bam"] = inp
applyBQSR["intervals"] = self.intervals
applyBQSR["reference"] = self.reference
self.step(
"mutect2",
gatk4.GatkMutect2_4_1_3(
normalBams=self.apply_bqsr_normal.out,
tumorBams=self.apply_bqsr_tumor.out,
normalSample=self.normal_name,
# tumorName=self.tumorName,
intervals=self.intervals,
reference=self.reference,
),
)
self.step(
"split_multi_allele",
SplitMultiAllele(reference=self.reference, vcf=self.mutect2.out),
)
self.output("out", source=self.split_multi_allele.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 constructor(self):
self.input("normal_bam", BamBai)
self.input("tumor_bam", BamBai)
self.input("normal_name", str)
self.input("tumor_name", str)
self.input(
"intervals",
Bed(),
doc="If this file 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.input("snps_1000gp", VcfTabix)
self.input("known_indels", VcfTabix)
self.input("mills_indels", VcfTabix)
self.input("gnomad", VcfTabix)
# # base calibration for normal and tumor bam
self.step(
"normal",
self.process_subpipeline(
bam=self.normal_bam,
intervals=self.targeted_bed,
reference=self.reference,
known_sites=[
self.snps_dbsnp,
self.snps_1000gp,
self.known_indels,
self.mills_indels,
],
),
)
self.step(
"tumor",
self.process_subpipeline(
bam=self.tumor_bam,
intervals=self.targeted_bed,
reference=self.reference,
known_sites=[
self.snps_dbsnp,
self.snps_1000gp,
self.known_indels,
self.mills_indels,
],
),
)
# variant calling + learn read orientation model
self.step(
"mutect2",
gatk4.GatkMutect2_4_1_3(
normalBams=self.normal.out,
tumorBams=self.tumor.out,
normalSample=self.normal_name,
intervals=self.targeted_bed,
reference=self.reference,
germlineResource=self.gnomad,
),
)
self.step(
"learnorientationmodel",
gatk4.Gatk4LearnReadOrientationModelLatest(
f1r2CountsFiles=self.mutect2.f1f2r_out, ),
)
# calculate contamination and segmentation
self.step(
"getpileupsummaries",
gatk4.Gatk4GetPileUpSummariesLatest(bam=self.tumor_bam,
sites=self.gnomad,
intervals=self.targeted_bed),
)
self.step(
"calculatecontamination",
gatk4.Gatk4CalculateContaminationLatest(
pileupTable=self.getpileupsummaries.out,
segmentationFileOut="tumor_segmentation.mutect2_segments",
),
)
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(
"splitnormalisevcf",
SplitMultiAlleleNormaliseVcf(
compressedTabixVcf=self.filtermutect2calls.out,
reference=self.reference),
)
self.step(
"filterpass",
VcfToolsvcftoolsLatest(
compressedVcf=self.splitnormalisevcf.out,
removeFileteredAll=True,
recode=True,
recodeINFOAll=True,
),
)
self.step("tabixvcf", TabixLatest(inp=self.filterpass.out))
self.output("variants", source=self.mutect2.out)
self.output("out", source=self.tabixvcf.out)