zzh
/
gatk-3.8
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

fixed a bug when the pipeline was used on a single bam. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5708 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
carneiro 2011-04-28 17:19:22 +00:00
parent b7334dcc1e
commit f04cc4321f
1 changed files with 6 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
scala/qscript/oneoffs/carneiro/dataProcessingV2.scala
View File

@ -153,9 +153,8 @@ class dataProcessingV2 extends QScript {
} }
// Checks how many contigs are in the dataset. Uses the BAM file header information. // Checks how many contigs are in the dataset. Uses the BAM file header information.
def getNumberOfContigs(): Int = { def getNumberOfContigs(bamFile: File): Int = {
val bam = fromFile(input).getLines.next val samReader = new SAMFileReader(new File(bamFile))
val samReader = new SAMFileReader(new File(bam))
return samReader.getFileHeader.getSequenceDictionary.getSequences.size() return samReader.getFileHeader.getSequenceDictionary.getSequences.size()
} }
@ -190,8 +189,8 @@ class dataProcessingV2 extends QScript {
// Generate a BAM file per sample joining all per lane files if necessary // Generate a BAM file per sample joining all per lane files if necessary
val sampleBamFiles = createSampleFiles(perLaneAlignedBamFiles) val sampleBamFiles = createSampleFiles(perLaneAlignedBamFiles)
// keep a record of the number of contigs in this bam file (they should all be the same // keep a record of the number of contigs in the first bam file in the list
nContigs = getNumberOfContigs() nContigs = getNumberOfContigs(perLaneAlignedBamFiles(0))
// Final output list of processed bam files // Final output list of processed bam files
var cohortList: List[File] = List() var cohortList: List[File] = List()
@ -288,19 +287,16 @@ class dataProcessingV2 extends QScript {
this.jobName = queueLogDir + outBam + ".clean" this.jobName = queueLogDir + outBam + ".clean"
} }
//todo -- add scatter gather capability (waiting for khalid's modifications to the queue base
case class cov (inBam: File, outRecalFile: File) extends CountCovariates with CommandLineGATKArgs { case class cov (inBam: File, outRecalFile: File) extends CountCovariates with CommandLineGATKArgs {
this.rodBind :+= RodBind("dbsnp", "VCF", dbSNP) this.rodBind :+= RodBind("dbsnp", "VCF", dbSNP)
this.covariate ++= List("ReadGroupCovariate", "QualityScoreCovariate", "CycleCovariate", "DinucCovariate") this.covariate ++= List("ReadGroupCovariate", "QualityScoreCovariate", "CycleCovariate", "DinucCovariate")
this.input_file :+= inBam this.input_file :+= inBam
this.recal_file = outRecalFile this.recal_file = outRecalFile
this.scatterCount = nContigs this.scatterCount = nContigs
this.jobQueue = "gsa" // should take this out once scatter gather is available.
this.analysisName = queueLogDir + outRecalFile + ".covariates" this.analysisName = queueLogDir + outRecalFile + ".covariates"
this.jobName = queueLogDir + outRecalFile + ".covariates" this.jobName = queueLogDir + outRecalFile + ".covariates"
} }
//todo -- add scatter gather capability (waiting for khalid's modifications to the queue base
case class recal (inBam: File, inRecalFile: File, outBam: File) extends TableRecalibration with CommandLineGATKArgs { case class recal (inBam: File, inRecalFile: File, outBam: File) extends TableRecalibration with CommandLineGATKArgs {
this.input_file :+= inBam this.input_file :+= inBam
this.recal_file = inRecalFile this.recal_file = inRecalFile
@ -309,10 +305,9 @@ class dataProcessingV2 extends QScript {
if (!qscript.intervalString.isEmpty()) this.intervalsString ++= List(qscript.intervalString) if (!qscript.intervalString.isEmpty()) this.intervalsString ++= List(qscript.intervalString)
else if (qscript.intervals != null) this.intervals :+= qscript.intervals else if (qscript.intervals != null) this.intervals :+= qscript.intervals
this.U = org.broadinstitute.sting.gatk.arguments.ValidationExclusion.TYPE.NO_READ_ORDER_VERIFICATION // todo -- update this with the last consensus between Tim, Matt and Eric. This is ugly! this.U = org.broadinstitute.sting.gatk.arguments.ValidationExclusion.TYPE.NO_READ_ORDER_VERIFICATION // todo -- update this with the last consensus between Tim, Matt and Eric. This is ugly!
this.index_output_bam_on_the_fly = true // todo -- implemente @gather for BAM index // this.index_output_bam_on_the_fly = true
// this.scatterCount = nContigs this.scatterCount = nContigs
this.isIntermediate = false this.isIntermediate = false
this.jobQueue = "gsa" // should take this out once scatter gather is available.
this.analysisName = queueLogDir + outBam + ".recalibration" this.analysisName = queueLogDir + outBam + ".recalibration"
this.jobName = queueLogDir + outBam + ".recalibration" this.jobName = queueLogDir + outBam + ".recalibration"