def constructor(self):
# Inputs
self.input("sample_name", String)
self.input("fastqs", Array(FastqGzPair))
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("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)
# 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
),
)
# 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 script here
# self.step("gridss_post_r", )
# 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,
),
)
# haploytype caller
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
),
)
self.step(
"addbamstats",
AddBamStatsGermline_0_1_0(
bam=self.merge_and_mark.out,
vcf=self.splitnormalisevcf.out,
reference=self.reference,
),
)
# 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", 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("hap_vcf", source=self.haplotype_caller.out, output_folder="VCF")
self.output("hap_bam", source=self.haplotype_caller.bam, output_folder="VCF")
self.output("normalise_vcf", source=self.addbamstats.out, output_folder="VCF")
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.input("snps_1000gp", VcfTabix)
self.input("known_indels", VcfTabix)
self.input("mills_indels", VcfTabix)
# self.step(
# "split_bam",
# gatk4.Gatk4SplitReads_4_0(bam=self.bam, intervals=self.intervals),
# )
self.step(
"base_recalibrator",
gatk4.Gatk4BaseRecalibrator_4_0(
bam=self.bam,
intervals=self.intervals,
reference=self.reference,
knownSites=[
self.snps_dbsnp,
self.snps_1000gp,
self.known_indels,
self.mills_indels,
],
),
)
self.step(
"apply_bqsr",
gatk4.Gatk4ApplyBqsr_4_0(
bam=self.bam,
intervals=self.intervals,
recalFile=self.base_recalibrator.out,
reference=self.reference,
),
)
self.step(
"haplotype_caller",
gatk4.Gatk4HaplotypeCaller_4_0(
inputRead=self.apply_bqsr,
intervals=self.intervals,
reference=self.reference,
dbsnp=self.snps_dbsnp,
pairHmmImplementation="LOGLESS_CACHING",
),
)
self.step(
"splitnormalisevcf",
SplitMultiAlleleNormaliseVcf(
compressedVcf=self.haplotype_caller.out,
reference=self.reference),
)
self.step(
"addbamstats",
AddBamStatsGermline_0_1_0(bam=self.bam,
vcf=self.splitnormalisevcf.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.addbamstats.out)
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):
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("gnomad", VcfTabix())
self.input("panel_of_normals", VcfTabix())
self.input("gatk_bam_str", String(optional=True))
# variant calling + learn read orientation model
self.step(
"mutect2",
gatk4.GatkMutect2_4_1_2(
tumorBams=self.bam,
intervals=self.intervals,
reference=self.reference,
panelOfNormals=self.panel_of_normals,
germlineResource=self.gnomad,
outputBamName=self.gatk_bam_str,
),
)
self.step(
"learnorientationmodel",
gatk4.Gatk4LearnReadOrientationModel_4_1_2(
f1r2CountsFiles=self.mutect2.f1f2r_out),
)
# calculate contamination and segmentation
self.step(
"getpileupsummaries",
gatk4.Gatk4GetPileUpSummaries_4_1_2(bam=self.bam,
sites=self.gnomad,
intervals=self.intervals),
)
self.step(
"calculatecontamination",
gatk4.Gatk4CalculateContamination_4_1_2(
pileupTable=self.getpileupsummaries.out),
)
self.step(
"filtermutect2calls",
gatk4.Gatk4FilterMutectCalls_4_1_2(
vcf=self.mutect2.out,
reference=self.reference,
segmentationFile=self.calculatecontamination.segOut,
contaminationTable=self.calculatecontamination.contOut,
readOrientationModel=self.learnorientationmodel.out,
statsFile=self.mutect2.stats,
),
)
# normalise vcf
self.step(
"splitnormalisevcf",
SplitMultiAlleleNormaliseVcf(
compressedTabixVcf=self.filtermutect2calls.out,
reference=self.reference,
),
)
self.output("variants", source=self.mutect2.out)
self.output("out_bam", source=self.mutect2.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", 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)