Merge branch 'patch-picard_summary' into 'develop'

Patch picard summary

changing some summaries for picard metrics

See merge request !163

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/core/summary/Summarizable.scala

@@ -30,7 +30,7 @@ trait Summarizable {
   def summaryFiles: Map[String, File]
 
   /** Must returns stats to store into summary */
-  def summaryStats: Map[String, Any]
+  def summaryStats: Any
 
   /** Can be used to add additional Summarizable, this is executed at the start of WriteSummary*/
   def addToQscriptSummary(qscript: SummaryQScript, name: String) {}

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/core/summary/WriteSummary.scala

@@ -139,7 +139,7 @@ class WriteSummary(val root: Configurable) extends InProcessFunction with Config
     val stats = summarizable.summaryStats
     val files = parseFiles(summarizable.summaryFiles)
 
-    (if (stats.isEmpty) Map[String, Any]() else Map("stats" -> Map(name -> stats))) ++
+    (Map("stats" -> Map(name -> stats))) ++
       (if (files.isEmpty) Map[String, Any]() else Map("files" -> Map(name -> files)))
   }
 

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CalculateHsMetrics.scala

@@ -63,7 +63,7 @@ class CalculateHsMetrics(val root: Configurable) extends Picard with Summarizabl
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats: Any = Picard.getMetrics(output).getOrElse(Map())
 }
 
 object CalculateHsMetrics {

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectAlignmentSummaryMetrics.scala

@@ -66,7 +66,7 @@ class CollectAlignmentSummaryMetrics(val root: Configurable) extends Picard with
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
 }
 
 object CollectAlignmentSummaryMetrics {

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectGcBiasMetrics.scala

@@ -71,7 +71,7 @@ class CollectGcBiasMetrics(val root: Configurable) extends Picard with Summariza
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getHistogram(output, tag = "METRICS CLASS")
 }
 
 object CollectGcBiasMetrics {

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectInsertSizeMetrics.scala

@@ -75,7 +75,7 @@ class CollectInsertSizeMetrics(val root: Configurable) extends Picard with Summa
   /** Returns files for summary */
   def summaryFiles: Map[String, File] = Map("output_histogram" -> outputHistogram)
 
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
 }
 
 object CollectInsertSizeMetrics {

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectMultipleMetrics.scala

@@ -68,25 +68,25 @@ class CollectMultipleMetrics(val root: Configurable) extends Picard with Summari
 
   override def addToQscriptSummary(qscript: SummaryQScript, name: String): Unit = {
     program.foreach(p => {
-      val stats: Map[String, Any] = p match {
+      val stats: Any = p match {
         case _ if p == Programs.CollectAlignmentSummaryMetrics.toString =>
-          Picard.getMetrics(new File(outputName + ".alignment_summary_metrics")).getOrElse(Map())
+          Picard.getMetrics(new File(outputName + ".alignment_summary_metrics"), groupBy = Some("CATEGORY"))
         case _ if p == Programs.CollectInsertSizeMetrics.toString =>
           Map(
-            "metrics" -> Picard.getMetrics(new File(outputName + ".insert_size_metrics")).getOrElse(Map()),
-            "histogram" -> Picard.getMetrics(new File(outputName + ".insert_size_metrics"), "HISTOGRAM").getOrElse(Map())
+            "metrics" -> Picard.getMetrics(new File(outputName + ".insert_size_metrics")),
+            "histogram" -> Picard.getHistogram(new File(outputName + ".insert_size_metrics"))
           )
         case _ if p == Programs.QualityScoreDistribution.toString =>
-          Picard.getMetrics(new File(outputName + ".quality_distribution_metrics"), "HISTOGRAM").getOrElse(Map())
+          Picard.getHistogram(new File(outputName + ".quality_distribution_metrics"))
         case _ if p == Programs.MeanQualityByCycle.toString =>
-          Picard.getMetrics(new File(outputName + ".quality_by_cycle_metrics"), "HISTOGRAM").getOrElse(Map())
+          Picard.getHistogram(new File(outputName + ".quality_by_cycle_metrics"))
         case _ if p == Programs.CollectBaseDistributionByCycle.toString =>
-          Picard.getMetrics(new File(outputName + ".base_distribution_by_cycle_metrics")).getOrElse(Map())
-        case _ => Map()
+          Picard.getHistogram(new File(outputName + ".base_distribution_by_cycle_metrics"), tag = "METRICS CLASS")
+        case _ => None
       }
       val sum = new Summarizable {
         override def summaryFiles: Map[String, File] = Map()
-        override def summaryStats: Map[String, Any] = stats
+        override def summaryStats = stats
       }
       qscript.addSummarizable(sum, p)
     })

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectRnaSeqMetrics.scala

@@ -84,7 +84,7 @@ class CollectRnaSeqMetrics(val root: Configurable) extends Picard with Summariza
       "output_chart" -> chartOutput
     ).collect { case (key, Some(value)) => key -> value }
 
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
 
   override def commandLine = super.commandLine +
     required("INPUT=", input, spaceSeparated = false) +

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectTargetedPcrMetrics.scala

@@ -51,7 +51,7 @@ class CollectTargetedPcrMetrics(val root: Configurable) extends Picard with Summ
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
 }
 
 object CollectTargetedPcrMetrics {

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/CollectWgsMetrics.scala

@@ -51,5 +51,5 @@ class CollectWgsMetrics(val root: Configurable) extends Picard with Summarizable
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
 }

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/MarkDuplicates.scala

@@ -100,7 +100,7 @@ class MarkDuplicates(val root: Configurable) extends Picard with Summarizable {
   def summaryFiles: Map[String, File] = Map()
 
   /** Returns stats for summary */
-  def summaryStats: Map[String, Any] = Picard.getMetrics(outputMetrics).getOrElse(Map())
+  def summaryStats = Picard.getMetrics(outputMetrics).getOrElse(Map())
 }
 object MarkDuplicates {
   /** Returns default MarkDuplicates */

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/extensions/picard/Picard.scala

@@ -20,7 +20,7 @@ import scala.io.Source
 
 import org.broadinstitute.gatk.utils.commandline.Argument
 
-import nl.lumc.sasc.biopet.core.BiopetJavaCommandLineFunction
+import nl.lumc.sasc.biopet.core.{ Logging, BiopetJavaCommandLineFunction }
 import nl.lumc.sasc.biopet.utils.tryToParseNumber
 
 /**
@@ -74,14 +74,51 @@ abstract class Picard extends BiopetJavaCommandLineFunction {
     conditional(createMd5, "CREATE_MD5_FILE=TRUE")
 }
 
-object Picard {
+object Picard extends Logging {
+
+  def getMetrics(file: File, tag: String = "METRICS CLASS",
+                 groupBy: Option[String] = None): Option[Any] = {
+    getMetricsContent(file, tag) match {
+      case Some((header, content)) => {
+        (content.size, groupBy) match {
+          case (_, Some(group)) => {
+            val groupId = header.indexOf(group)
+            if (groupId == -1) throw new IllegalArgumentException(group + " not existing in header of: " + file)
+            if (header.count(_ == group) > 1) logger.warn(group + " multiple times seen in header of: " + file)
+            Some((for (c <- content) yield c(groupId).toString() -> {
+              header.filter(_ != group).zip(c.take(groupId) ::: c.takeRight(c.size - groupId - 1)).toMap
+            }).toMap)
+          }
+          case (1, _) => Some(header.zip(content.head).toMap)
+          case _      => Some(header :: content)
+        }
+      }
+      case _ => None
+    }
+  }
+
+  /**
+   * This function parse the metrics but transpose for table
+   * @param file metrics file
+   * @param tag default to "HISTOGRAM"
+   * @return
+   */
+  def getHistogram(file: File, tag: String = "HISTOGRAM") = {
+    getMetricsContent(file, tag) match {
+      case Some((header, content)) => {
+        val colums = header.zipWithIndex.map(x => x._1 -> content.map(_.lift(x._2))).toMap
+        Some(colums)
+      }
+      case _ => None
+    }
+  }
 
   /**
    * This function parse a metrics file in separated values
    * @param file input metrics file
    * @return (header, content)
    */
-  def getMetrics(file: File, tag: String = "METRICS CLASS"): Option[Map[String, Any]] =
+  def getMetricsContent(file: File, tag: String) = {
     if (!file.exists) None
     else {
       val lines = Source.fromFile(file).getLines().toArray
@@ -94,6 +131,7 @@ object Picard {
         lines(i).split("\t").map(v => tryToParseNumber(v, true).getOrElse(v)).toList
       }).toList
 
-      Some(Map("content" -> (header :: content)))
+      Some(header, content)
     }
+  }
 }
\ No newline at end of file

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/utils/ConfigUtils.scala

@@ -160,6 +160,7 @@ object ConfigUtils extends Logging {
     any match {
       case j: Json      => j
       case None         => Json.jNull
+      case Some(x)      => anyToJson(x)
       case m: Map[_, _] => mapToJson(m.map(m => m._1.toString -> anyToJson(m._2)))
       case l: List[_]   => Json.array(l.map(anyToJson(_)): _*)
       case b: Boolean   => Json.jBool(b)