def mapEachInterval(self, workflow=None, intervalData=None, chromosome=None, \
VCFJobData=None, passingData=None,
mapEachChromosomeData=None, transferOutput=False, \
**keywords):
"""
2013.04.30
"""
if workflow is None:
workflow = self
returnData = PassingData(no_of_jobs = 0)
returnData.jobDataLs = []
topOutputDirJob = passingData.topOutputDirJob
intervalFileBasenamePrefix = passingData.intervalFileBasenamePrefix
span = passingData.span
noOfIndividuals= passingData.noOfIndividuals
SNPVCFFile = VCFJobData.file
SNPVCFJobLs = VCFJobData.jobLs
"""
### 2013.06.19 intervalData does not exsit for input that is entirely composed of VCF files (SplitVCFFile job does not return intervals)
if intervalData.file:
mpileupInterval = intervalData.interval
bcftoolsInterval = intervalData.file
else:
mpileupInterval = intervalData.interval
bcftoolsInterval = intervalData.interval
intervalFileBasenameSignature = intervalData.intervalFileBasenameSignature
overlapInterval = intervalData.overlapInterval
overlapFileBasenameSignature = intervalData.overlapIntervalFnameSignature
span = intervalData.span
"""
if chromosome is None:
chromosome = getattr(passingData, 'chromosome', None)
#noOfIndividuals
realInputVolume = noOfIndividuals * span
baseInputVolume = 600*2000 #600 individuals at 2000 sites
walltime = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=10000).value
#splitVCFJob = passingData.mapEachVCFData.splitVCFJob
#### Part 1 generate high-quality reference panel through Beagle on high-coverage individuals
# extractRefPanelSampleIDJob outputs sample IDs with replicate tags
# select the high-coverage members
outputVCF = File(os.path.join(self.highCoveragePanelDirJob.output, \
'%s.minCoverageForRefPanel%s.vcf'%(intervalFileBasenamePrefix, self.minCoverageForRefPanel)))
#selectVariants would re-generate AC, AF so that TrioCaller could read it.
#samtools uses 'AC1' instead of AC, 'AF1' instead of AF.
selectHighCoverageSampleJob = self.addSelectVariantsJob(SelectVariantsJava=self.SelectVariantsJava, \
inputF=VCFJobData.file, outputF=outputVCF, \
refFastaFList=self.registerReferenceData.refFastaFList, sampleIDKeepFile=self.extractRefPanelSampleIDJob.output,\
parentJobLs=[self.highCoveragePanelDirJob, self.extractRefPanelSampleIDJob] + VCFJobData.jobLs, \
extraDependentInputLs=[VCFJobData.tbi_F], transferOutput=False, \
extraArguments=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=9000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value)
# run Beagle
outputFnamePrefix = os.path.join(self.highCoveragePanelDirJob.folder, "%s.minCoverage%s.beagled"%\
(intervalFileBasenamePrefix, self.minCoverageForRefPanel))
beagleOnHighCoverageJob = self.addBeagle4Job(executable=self.BeagleOnHCMOnkeys, \
inputFile=selectHighCoverageSampleJob.output, refPanelFile=None,\
pedFile = self.outputPedigreeOfHghCoverageSamplesJob.output,\
outputFnamePrefix=outputFnamePrefix, \
burninIterations=7, phaseIterations=10, \
noOfSamplingHaplotypesPerSample=4,\
parentJobLs=[self.highCoveragePanelDirJob, selectHighCoverageSampleJob, self.outputPedigreeOfHghCoverageSamplesJob], \
transferOutput=False, \
extraArguments=None, extraArgumentList=None,\
extraOutputLs=None, extraDependentInputLs=None, \
no_of_cpus=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=13000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value)
#index .vcf.gz, output of beagle, without index, GATK can't work on gzipped vcf
tabixIndexFile = File('%s.tbi'%(beagleOnHighCoverageJob.output.name))
tabixOnHighCoverageVCFJob = self.addGenericJob(executable=self.tabix, \
inputFile=beagleOnHighCoverageJob.output, inputArgumentOption="",\
outputFile=None, outputArgumentOption="-o", \
extraDependentInputLs=None, \
extraOutputLs=[tabixIndexFile], transferOutput=False, frontArgumentList=["-p vcf"], \
extraArguments=None, \
extraArgumentList=[], \
parentJobLs=[beagleOnHighCoverageJob, self.highCoveragePanelDirJob],\
no_of_cpus=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=2000, maxJobPropertyValue=5000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=600).value)
# select the high-coverage members
outputVCF = File(os.path.join(self.highCoveragePanelDirJob.output, \
'%s.minCoverage%s.maxPairwiseKinship%s.refPanel.beagled.vcf'%\
(intervalFileBasenamePrefix, self.minCoverageForRefPanel, self.maxPairwiseKinship)))
#selectVariants would re-generate AC, AF so that TrioCaller could read it.
#samtools uses 'AC1' instead of AC, 'AF1' instead of AF.
selectDistantMembersVariantsJob = self.addSelectVariantsJob(SelectVariantsJava=self.SelectVariantsJava, \
inputF=beagleOnHighCoverageJob.output, outputF=outputVCF, \
refFastaFList=self.registerReferenceData.refFastaFList, \
sampleIDKeepFile=self.selectDistantMembersFromGenotypeFileJob.output,\
parentJobLs=[self.highCoveragePanelDirJob, beagleOnHighCoverageJob, self.selectDistantMembersFromGenotypeFileJob,\
tabixOnHighCoverageVCFJob], \
extraDependentInputLs=[tabixOnHighCoverageVCFJob.output], transferOutput=False, \
extraArguments=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=7000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value)
##### Part 2 run Beagle on everyone with reference panel
# run Beagle
#refPanelFile=selectDistantMembersVariantsJob.output,\
outputFnamePrefix = os.path.join(self.mapDirJob.folder, '%s.beagled'%(intervalFileBasenamePrefix))
beagleJob = self.addBeagle4Job(executable=self.BeagleJava, \
inputFile=VCFJobData.file, refPanelFile=None,\
pedFile=self.outputPedigreeJob.output,\
outputFnamePrefix=outputFnamePrefix, \
burninIterations=7, phaseIterations=10, \
noOfSamplingHaplotypesPerSample=4, duoscale=2, trioscale=2, \
extraArguments=None, extraArgumentList=None,\
parentJobLs=[self.mapDirJob, \
self.outputPedigreeJob] + VCFJobData.jobLs, \
transferOutput=False, no_of_cpus=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=13000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value,\
)
returnData.beagleJob = beagleJob
#index .vcf.gz, output of beagle, without index, GATK can't work on gzipped vcf
tabixIndexFile = File('%s.tbi'%(beagleJob.output.name))
tabixJob = self.addGenericJob(executable=self.tabix, \
inputFile=beagleJob.output, inputArgumentOption="",\
outputFile=None, outputArgumentOption="-o", \
extraDependentInputLs=None, \
extraOutputLs=[tabixIndexFile], transferOutput=False, frontArgumentList=["-p vcf"], \
extraArguments=None, \
extraArgumentList=[], \
parentJobLs=[beagleJob, self.mapDirJob],\
no_of_cpus=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=2000, maxJobPropertyValue=4000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=600).value)
#borrow PL to from pre-Beagle VCF to genotype
outputFile = File(os.path.join(self.mapDirJob.folder, '%s.beagled.withPL.vcf'%(intervalFileBasenamePrefix)))
combineBeagleAndPreBeagleVariantsJob = self.addGATKJob(executable=self.CombineBeagleAndPreBeagleVariantsJava, \
GenomeAnalysisTKJar=self.GenomeAnalysisTKJar, \
GATKAnalysisType="CombineBeagleAndPreBeagleVariants",\
inputFile=None, inputArgumentOption=None, \
refFastaFList=self.registerReferenceData.refFastaFList, \
inputFileList=None, argumentForEachFileInInputFileList="--variant",\
interval=None, outputFile=outputFile, outputArgumentOption="--out", \
frontArgumentList=None, extraArguments=None, \
extraArgumentList=["--variant:first", beagleJob.output, "--variant:second", VCFJobData.file, \
"-genotypeMergeOptions PRIORITIZE", "-priority first,second"], \
extraOutputLs=None, \
extraDependentInputLs=[beagleJob.output, VCFJobData.file] + tabixJob.outputLs, \
parentJobLs=[beagleJob, tabixJob]+ VCFJobData.jobLs, transferOutput=False, \
no_of_cpus=None, \
key2ObjectForJob=None,\
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=2000, maxJobPropertyValue=4000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=600).value)
#do not use "--variant:beagle" to name your vcf file as GATK would think it's in Beagle format
#TrioCaller
# 2013.06.11 replicate individuals who appear in more than 1 families
round1_IndividualsReplicatedVCF = File( os.path.join(self.mapDirJob.folder, \
'%s.replicate.vcf'%(intervalFileBasenamePrefix)))
replicateVCFGenotypeColumnsJob = self.addReplicateVCFGenotypeColumnsJob(\
executable=self.ReplicateVCFGenotypeColumns, \
inputF=combineBeagleAndPreBeagleVariantsJob.output, \
sampleID2FamilyCountF=self.outputReplicatePedigreeJob.sampleID2FamilyCountF, \
outputF=round1_IndividualsReplicatedVCF, \
replicateIndividualTag=self.replicateIndividualTag,\
parentJobLs=[self.outputReplicatePedigreeJob, self.mapDirJob, combineBeagleAndPreBeagleVariantsJob], \
extraDependentInputLs=None, \
transferOutput=False, \
extraArguments=None, \
job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=9000).value, \
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value,\
)
refineGenotypeOutputF = File(os.path.join(self.mapDirJob.folder, \
'%s.trioCaller.vcf'%(intervalFileBasenamePrefix)))
refineGenotypeJob = self.addTrioCallerJob(trioCallerWrapper=self.trioCallerWrapper, \
trioCallerPath=self.trioCallerPath, \
inputVCF=replicateVCFGenotypeColumnsJob.output,\
pedFile=self.outputReplicatePedigreeJob.output, outputVCF=refineGenotypeOutputF, \
inputPhased=True,\
parentJobLs=[self.mapDirJob, replicateVCFGenotypeColumnsJob, self.outputReplicatePedigreeJob], \
extraDependentInputLs=[], transferOutput=False, \
extraArguments=None, \
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=4000, maxJobPropertyValue=9000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value) #1.2G memory for 12K loci
returnData.refineGenotypeJob = refineGenotypeJob
"""
2013.07.10 the TrioCaller VCF has some info tags that are not described in VCF header
"""
outputFile = File(os.path.join(self.mapDirJob.folder, \
'%s.extraInfoDesc.vcf'%(intervalFileBasenamePrefix)))
addInfoDescJob = self.addGenericJob(executable=self.AddMissingInfoDescriptionToVCFHeader, \
inputFile=refineGenotypeJob.output, \
inputArgumentOption="-i", \
outputFile=outputFile, outputArgumentOption="-o", \
parentJobLs=[self.mapDirJob, refineGenotypeJob], \
extraDependentInputLs=None, extraOutputLs=None, \
frontArgumentList=None, extraArguments=None, extraArgumentList=None, \
transferOutput=False, sshDBTunnel=None, \
key2ObjectForJob=None, objectWithDBArguments=None, \
no_of_cpus=None,
job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=2000, \
minJobPropertyValue=1000, maxJobPropertyValue=3000).value, \
walltime=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=500).value,\
max_walltime=None)
# a CheckGenotypeConcordanceAmongReplicates.py job
trioCallerReplicateConcordanceFile = File(os.path.join(self.statDirJob.folder, \
'%s.trioCaller.concordance.tsv'%(intervalFileBasenamePrefix)))
returnData.trioCallerReplicateConcordanceJob = self.addGATKJob(executable=self.CalculateConcordanceJava, \
GenomeAnalysisTKJar=self.GenomeAnalysisTKJar, \
GATKAnalysisType="CalculateConcordanceAmongReplicates",\
inputFile=refineGenotypeJob.output, inputArgumentOption="--variant", \
refFastaFList=self.registerReferenceData.refFastaFList, \
interval=None, \
outputFile=trioCallerReplicateConcordanceFile, outputArgumentOption="--concordanceStatFname",\
frontArgumentList=None, extraArguments="--replicateIndividualTag %s"%(self.replicateIndividualTag), \
extraArgumentList=None, extraOutputLs=None, \
parentJobLs=[self.statDirJob, refineGenotypeJob], \
transferOutput=False, \
no_of_cpus=None, \
job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=6000, \
minJobPropertyValue=9000, maxJobPropertyValue=16000).value, \
walltime=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value)
#2013.06.14
#merge replicates to generate consensus call
# (not haplotype-based, as different recombination points across replicate haplotypes make it non-trivial )
mergeReplicateOutputF = File(os.path.join(self.mapDirJob.folder, \
'%s.replicatesMerged.vcf'%(intervalFileBasenamePrefix)))
returnData.mergeVCFReplicateColumnsJob = self.addMergeVCFReplicateGenotypeColumnsJob(\
executable=self.MergeVCFReplicateHaplotypesJava,\
GenomeAnalysisTKJar=self.GenomeAnalysisTKJar, \
inputF=addInfoDescJob.output, outputF=mergeReplicateOutputF, \
replicateIndividualTag=self.replicateIndividualTag, \
refFastaFList=self.registerReferenceData.refFastaFList, \
parentJobLs=[self.mapDirJob, addInfoDescJob], \
extraDependentInputLs=[], transferOutput=False, \
extraArguments=None, \
analysis_type='MergeVCFReplicateGenotypeColumns',\
job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \
minJobPropertyValue=5000, maxJobPropertyValue=9000).value,\
walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \
baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \
minJobPropertyValue=60, maxJobPropertyValue=1200).value)
return returnData
def addRefineGenotypeJobsViaTrioCaller(self, inputFile=None, vcfBaseFname=None, outputDirJob=None, statDirJob=None, \ refFastaFList=None, intervalData=None,\ baseInputVolume=450*2000000, realInputVolume=None,\ parentJobLs=None, \ transferOutput=False, \ no_of_cpus=None, job_max_memory=2000, walltime=180, \ max_walltime=None, **keywords): returnData = PassingData() if getattr(self, "outputPedigreeJob", None) is None: outputFileFormat = 2 #trioCaller pedFile = File(os.path.join(outputDirJob.output, 'pedigree_outputFileFormat%s.txt'%(outputFileFormat))) sampleID2FamilyCountF = File(os.path.join(outputDirJob.output, 'sampleID2FamilyCount_outputFileFormat%s.txt'%(outputFileFormat))) if outputFileFormat==3: #for polymutt polymuttDatFile = File(os.path.join(outputDirJob.output, 'datFile_outputFileFormat%s.txt'%(outputFileFormat))) else: polymuttDatFile = None self.outputPedigreeJob = self.addOutputVRCPedigreeInTFAMGivenOrderFromFileJob(executable=self.OutputVRCPedigreeInTFAMGivenOrderFromFile, \ inputFile=inputFile, \ outputFile=pedFile, sampleID2FamilyCountF=sampleID2FamilyCountF,\ polymuttDatFile = polymuttDatFile,\ outputFileFormat=outputFileFormat, replicateIndividualTag=self.replicateIndividualTag,\ treatEveryOneIndependent=False,\ parentJobLs=parentJobLs + [outputDirJob], extraDependentInputLs=None, transferOutput=True, \ extraArguments=None, job_max_memory=2000, sshDBTunnel=self.needSSHDBTunnel) SNPOnlyOutputF = File(os.path.join(outputDirJob.folder, '%s.SNP.vcf'%vcfBaseFname)) selectSNPJob = self.addGATKJob(executable=self.SelectVariantsJava, GATKAnalysisType="SelectVariants",\ inputFile=inputFile, inputArgumentOption="--variant", refFastaFList=refFastaFList, inputFileList=None,\ argumentForEachFileInInputFileList=None,\ interval=None, outputFile=SNPOnlyOutputF, \ parentJobLs=[outputDirJob] + parentJobLs, transferOutput=False, job_max_memory=job_max_memory,\ frontArgumentList=None, extraArguments="-selectType SNP", \ extraArgumentList=None, extraOutputLs=None, \ extraDependentInputLs=None, no_of_cpus=None, walltime=80) # 2013.06.11 replicate individuals who appear in more than 1 families round1_IndividualsReplicatedVCF = File( os.path.join(outputDirJob.folder, \ '%s.replicate.vcf'%(vcfBaseFname))) replicateVCFGenotypeColumnsJob = self.addReplicateVCFGenotypeColumnsJob(\ executable=self.ReplicateVCFGenotypeColumns, \ inputF=selectSNPJob.output, \ sampleID2FamilyCountF=self.outputPedigreeJob.sampleID2FamilyCountF, \ outputF=round1_IndividualsReplicatedVCF, \ replicateIndividualTag=self.replicateIndividualTag,\ parentJobLs=[self.outputPedigreeJob, outputDirJob, selectSNPJob], \ extraDependentInputLs=None, \ transferOutput=False, \ extraArguments=None, \ job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \ minJobPropertyValue=4000, maxJobPropertyValue=9000).value, \ walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \ minJobPropertyValue=60, maxJobPropertyValue=1200).value,\ ) trioCallerOutputFile = File(os.path.join(outputDirJob.folder, \ '%s.trioCaller.vcf'%(vcfBaseFname))) trioCallerJob = self.addTrioCallerJob(trioCallerWrapper=self.trioCallerWrapper, \ trioCallerPath=self.trioCallerPath, \ inputVCF=replicateVCFGenotypeColumnsJob.output,\ pedFile=self.outputPedigreeJob.output, outputVCF=trioCallerOutputFile, \ inputPhased=False,\ parentJobLs=[outputDirJob, replicateVCFGenotypeColumnsJob, self.outputPedigreeJob], \ extraDependentInputLs=[], transferOutput=False, \ extraArguments=None, \ job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \ minJobPropertyValue=4000, maxJobPropertyValue=9000).value,\ walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \ minJobPropertyValue=60, maxJobPropertyValue=1200).value) #1.2G memory for 12K loci returnData.trioCallerJob = trioCallerJob """ 2013.07.10 the TrioCaller VCF has some info tags that are not described in VCF header """ outputFile = File(os.path.join(outputDirJob.folder, \ '%s.extraInfoDesc.vcf'%(vcfBaseFname))) addInfoDescJob = self.addGenericJob(executable=self.AddMissingInfoDescriptionToVCFHeader, \ inputFile=trioCallerJob.output, \ inputArgumentOption="-i", \ outputFile=outputFile, outputArgumentOption="-o", \ parentJobLs=[outputDirJob, trioCallerJob], \ extraDependentInputLs=None, extraOutputLs=None, \ frontArgumentList=None, extraArguments=None, extraArgumentList=None, \ transferOutput=False, sshDBTunnel=None, \ key2ObjectForJob=None, objectWithDBArguments=None, \ no_of_cpus=None, job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=2000, \ minJobPropertyValue=1000, maxJobPropertyValue=3000).value, \ walltime=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \ minJobPropertyValue=60, maxJobPropertyValue=500).value,\ max_walltime=None) if statDirJob: # a CheckGenotypeConcordanceAmongReplicates.py job trioCallerReplicateConcordanceFile = File(os.path.join(self.statDirJob.folder, \ '%s.trioCaller.concordance.tsv'%(vcfBaseFname))) returnData.trioCallerReplicateConcordanceJob = self.addGATKJob(executable=self.CalculateConcordanceJava, \ GenomeAnalysisTKJar=self.GenomeAnalysisTKJar, \ GATKAnalysisType="CalculateConcordanceAmongReplicates",\ inputFile=trioCallerJob.output, inputArgumentOption="--variant", \ refFastaFList=self.registerReferenceData.refFastaFList, \ interval=None, \ outputFile=trioCallerReplicateConcordanceFile, outputArgumentOption="--concordanceStatFname",\ frontArgumentList=None, extraArguments="--replicateIndividualTag %s"%(self.replicateIndividualTag), \ extraArgumentList=None, extraOutputLs=None, \ parentJobLs=[self.statDirJob, trioCallerJob], \ transferOutput=False, \ no_of_cpus=None, \ job_max_memory=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=6000, \ minJobPropertyValue=9000, maxJobPropertyValue=16000).value, \ walltime=self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \ minJobPropertyValue=60, maxJobPropertyValue=1200).value) #2013.06.14 #merge replicates to generate consensus call # (not haplotype-based, as different recombination points across replicate haplotypes make it non-trivial ) mergeReplicateOutputF = File(os.path.join(outputDirJob.folder, \ '%s.replicatesMerged.vcf'%(vcfBaseFname))) mergeVCFReplicateColumnsJob = self.addMergeVCFReplicateGenotypeColumnsJob(\ executable=self.MergeVCFReplicateHaplotypesJava,\ GenomeAnalysisTKJar=self.GenomeAnalysisTKJar, \ inputF=addInfoDescJob.output, outputF=mergeReplicateOutputF, \ replicateIndividualTag=self.replicateIndividualTag, \ refFastaFList=self.registerReferenceData.refFastaFList, \ parentJobLs=[outputDirJob, addInfoDescJob], \ extraDependentInputLs=[], transferOutput=False, \ extraArguments=None, \ analysis_type='MergeVCFReplicateGenotypeColumns',\ job_max_memory = self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=4000, \ minJobPropertyValue=5000, maxJobPropertyValue=9000).value,\ walltime= self.scaleJobWalltimeOrMemoryBasedOnInput(realInputVolume=realInputVolume, \ baseInputVolume=baseInputVolume, baseJobPropertyValue=60, \ minJobPropertyValue=60, maxJobPropertyValue=1200).value) returnData.refineGenotypeJob = mergeVCFReplicateColumnsJob #the final gentoype job returnData.refineGenotypeJob.intervalData = intervalData #attached so that it could be used by downstream jobs return returnData
