Better DOT support and updated recalibration pipeline

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@3811 348d0f76-0448-11de-a6fe-93d51630548a

scala/qscript/recalibrate.scala

@@ -1,3 +1,4 @@
+import java.io.File
 import org.broadinstitute.sting.queue.QScript._
 import org.apache.commons.io.FilenameUtils;
 // Other imports can be added here
@@ -35,16 +36,22 @@ setParams
 // Run the pipeline
 run
 
+def bai(bam: File) = new File(bam + ".bai")
+
 class Index(bamIn: File) extends GatkFunction {
     @Input(doc="foo") var bam = bamIn
+    @Output(doc="foo") var bamIndex = bai(bamIn)
     memoryLimit = Some(1)
+    override def dotString = "Index: %s".format(bamIn.getName)
     def commandLine = "samtools index %s".format(bam)
 }
 
 class CountCovariates(bamIn: File, recalDataIn: File, args: String = "") extends GatkFunction {
     @Input(doc="foo") var bam = bamIn
+    @Input(doc="foo") var bamIndex = bai(bamIn)
     @Output(doc="foo") var recalData = recalDataIn
     memoryLimit = Some(4)
+    override def dotString = "CountCovariates: %s [args %s]".format(bamIn.getName, args)
     def commandLine = gatkCommandLine("CountCovariates") + args + " -l INFO -D /humgen/gsa-hpprojects/GATK/data/dbsnp_129_hg18.rod -I %s --max_reads_at_locus 20000 -cov ReadGroupCovariate -cov QualityScoreCovariate -cov CycleCovariate -cov DinucCovariate -recalFile %s".format(bam, recalData)
 }
 
@@ -53,6 +60,7 @@ class TableRecalibrate(bamInArg: File, recalDataIn: File, bamOutArg: File, args:
     @Input(doc="foo") var recalData = recalDataIn
     @Gather(classOf[BamGatherFunction])
     @Output(doc="foo") var bamOut = bamOutArg
+    override def dotString = "TableRecalibrate: %s => %s [args %s]".format(bamInArg.getName, bamOutArg.getName, args)
     memoryLimit = Some(2)
     def commandLine = gatkCommandLine("TableRecalibration") + args + " -l INFO -I %s -recalFile %s -outputBam %s".format(bamIn, recalData, bamOut) // bamOut.getPath())
 }
@@ -60,5 +68,6 @@ class TableRecalibrate(bamInArg: File, recalDataIn: File, bamOutArg: File, args:
 class AnalyzeCovariates(recalDataIn: File, outputDir: File) extends GatkFunction {
     @Input(doc="foo") var recalData = recalDataIn
     memoryLimit = Some(4)
+    override def dotString = "AnalyzeCovariates: %s".format(recalDataIn.getName)
     def commandLine = "java -Xmx4g -jar /home/radon01/depristo/dev/GenomeAnalysisTK/trunk/dist/AnalyzeCovariates.jar -recalFile %s -outputDir %s -resources /home/radon01/depristo/dev/GenomeAnalysisTK/trunk/R/ -ignoreQ 5 -Rscript /broad/tools/apps/R-2.6.0/bin/Rscript".format(recalData, outputDir)
 }

scala/src/org/broadinstitute/sting/queue/engine/QGraph.scala

@@ -157,10 +157,7 @@ class QGraph extends Logging {
     // todo -- we need a nice way to visualize the key pieces of information about commands.  Perhaps a
     // todo -- visualizeString() command, or something that shows inputs / outputs
     val ve = new org.jgrapht.ext.EdgeNameProvider[QFunction] {
-        def getEdgeName( function: QFunction ) = function match {
-            case f: DispatchFunction => f.jobName + " => " + f.commandLine
-            case _ => ""
-        }
+        def getEdgeName( function: QFunction ) = function.dotString
     }
 
     //val iterator = new TopologicalOrderIterator(qGraph.jobGraph)

scala/src/org/broadinstitute/sting/queue/function/DispatchFunction.scala

@@ -57,6 +57,8 @@ trait DispatchFunction extends InputOutputFunction {
     super.freeze
   }
 
+  override def dotString = jobName + " => " + commandLine
+
   /**
    * Override the canon function to change any relative path to an absolute path.
    */

scala/src/org/broadinstitute/sting/queue/function/QFunction.scala

@@ -22,4 +22,6 @@ trait QFunction {
    * Set of outputs for this function.
    */
   def outputs: Set[Any]
+
+  def dotString = ""
 }