Merge branch 'feature-gentrap_additions' into 'develop'

Feature gentrap additions

This is the latest updates and fixes on the Gentrap pipeline to be compatible with the most recent framework updates.

In particular:

* Issue #156 is fixed ~ this was due to class initialization problems
* Issue #157 is fixed
* TopHat alignment results are now merged with its unmapped alignment file, producing the correct statistics
* The pipeline is updated with the new reference module.
* Redundant CollectRnaSeqMetrics is removed (now using the one from BamMetrics instead).

See merge request !181

docs/examples/gentrap_example.json

+{
+  "samples" : {
+    "sampleA" : {
+      "libraries" : {
+        "lib_1" : {
+          "R1" : "/path/to/inputA_R1.fq.gz",
+          "R2" : "/path/to/inputA_R2.fq.gz"
+        }
+      }
+    },
+    "sampleB" : {
+      "libraries" : {
+        "lib_1" : {
+          "R1" : "/path/to/inputB_1_R1.fq.gz",
+          "R2" : "/path/to/inputB_1_R2.fq.gz"
+        },
+        "lib_2": {
+          "R1" : "/path/to/inputB_2_R1.fq.gz",
+          "R2" : "/path/to/inputB_2_R2.fq.gz"
+        }
+      }
+    }
+  },
+  "gentrap": {
+    "output_dir": "/path/to/output_dir",
+    "expression_measures": ["fragments_per_gene", "bases_per_gene", "bases_per_exon"],
+    "strand_protocol": "non_specific",
+    "aligner": "gsnap",
+    "reference": "/share/isilon/system/local/Genomes-new-27-10-2011/H.Sapiens/hg19_nohap/gsnap/reference.fa",
+    "annotation_gtf": "/path/to/data/annotation/ucsc_refseq.gtf",
+    "annotation_bed": "/path/to/data/annotation/ucsc_refseq.bed",
+    "annotation_refflat": "/path/to/data/annotation/ucsc_refseq.refFlat",
+    "gsnap": {
+      "dir": "/share/isilon/system/local/Genomes-new-27-10-2011/H.Sapiens/hg19_nohap/gsnap",
+      "db": "hg19_nohap",
+      "quiet_if_excessive": true,
+      "npaths": 1
+    },
+    "cutadapt": {
+      "minimum_length": 20
+    },
+    "mapping": {
+      "flexiprep": {
+        "fastqc": {
+          "threads": 6,
+          "nogroup": true
+        }
+      }
+    },
+    "rawbasecounter": {
+      "core_memory": "20G"
+    }
+  }
+}

docs/pipelines/gentrap.md

@@ -110,18 +110,23 @@ Thus, an example settings configuration is as follows:
 }
 ~~~
 
+#### Example configurations
+
+In most cases, it's practical to combine the samples and settings configuration into one file. Here is an [example config file](/examples/gentrap_example.json) where both samples and settings are stored into one file. Note also that there are additional tool configurations in the config file.
+
 ## Running Gentrap
 
 As with other pipelines in the Biopet suite, Gentrap can be run by specifying the pipeline after the `pipeline` subcommand:
 
 ~~~
-java -jar </path/to/biopet.jar> pipeline gentrap -config </path/to/config.json> -qsub -jobParaEnv BWA -run
+$ java -jar </path/to/biopet.jar> pipeline gentrap -config </path/to/config.json> -qsub -jobParaEnv BWA -run
 ~~~
 
-If you already have the `biopet` environment module loaded, you can also simply call `biopet`:
+You can also use the `biopet` environment module (recommended) when you are running the pipeline in SHARK:
 
 ~~~
-biopet pipeline gentrap -config </path/to/config.json> -qsub -jobParaEnv BWA -run
+$ module load biopet/v0.3.1
+$ biopet pipeline gentrap -config </path/to/config.json> -qsub -jobParaEnv BWA -run
 ~~~
 
 It is also a good idea to specify retries (we recomend `-retry 3` up to `-retry 5`) so that cluster glitches do not interfere with your pipeline runs.

public/bammetrics/src/main/scala/nl/lumc/sasc/biopet/pipelines/bammetrics/BamMetrics.scala

@@ -38,7 +38,9 @@ class BamMetrics(val root: Configurable) extends QScript with SummaryQScript wit
   /** Bed of amplicon that is used */
   var ampliconBedFile: Option[File] = config("amplicon_bed")
 
-  var rnaMetrics: Boolean = config("rna_metrcis", default = false)
+  /** Settings for CollectRnaSeqMetrics */
+  var rnaMetricsSettings: Map[String, String] = Map()
+  var transcriptRefFlatFile: Option[File] = config("transcript_refflat")
 
   /** return location of summary file */
   def summaryFile = (sampleId, libId) match {
@@ -78,17 +80,22 @@ class BamMetrics(val root: Configurable) extends QScript with SummaryQScript wit
     add(gcBiasMetrics)
     addSummarizable(gcBiasMetrics, "gc_bias")
 
-    val wgsMetrics = new CollectWgsMetrics(this)
-    wgsMetrics.input = inputBam
-    wgsMetrics.output = swapExt(outputDir, inputBam, ".bam", ".wgs.metrics")
-    add(wgsMetrics)
-    addSummarizable(wgsMetrics, "wgs")
+    if (transcriptRefFlatFile.isEmpty) {
+      val wgsMetrics = new CollectWgsMetrics(this)
+      wgsMetrics.input = inputBam
+      wgsMetrics.output = swapExt(outputDir, inputBam, ".bam", ".wgs.metrics")
+      add(wgsMetrics)
+      addSummarizable(wgsMetrics, "wgs")
+    }
 
-    if (rnaMetrics) {
+    if (transcriptRefFlatFile.isDefined) {
       val rnaMetrics = new CollectRnaSeqMetrics(this)
       rnaMetrics.input = inputBam
       rnaMetrics.output = swapExt(outputDir, inputBam, ".bam", ".rna.metrics")
       rnaMetrics.chartOutput = Some(swapExt(outputDir, inputBam, ".bam", ".rna.metrics.pdf"))
+      rnaMetrics.refFlat = transcriptRefFlatFile.get
+      rnaMetrics.ribosomalIntervals = rnaMetricsSettings.get("ribosomal_intervals").collect { case n => new File(n) }
+      rnaMetrics.strandSpecificity = rnaMetricsSettings.get("strand_specificity")
       add(rnaMetrics)
       addSummarizable(rnaMetrics, "rna")
     }

public/bammetrics/src/test/scala/nl/lumc/sasc/biopet/pipelines/bammetrics/BamMetricsTest.scala

@@ -45,9 +45,10 @@ class BamMetricsTest extends TestNGSuite with Matchers {
 
   @Test(dataProvider = "bammetricsOptions")
   def testFlexiprep(rois: Int, amplicon: Boolean, rna: Boolean) = {
-    val map = ConfigUtils.mergeMaps(Map("output_dir" -> BamMetricsTest.outputDir, "rna_metrcis" -> rna
-    ), Map(BamMetricsTest.executables.toSeq: _*)) ++
+    val map = ConfigUtils.mergeMaps(Map("output_dir" -> BamMetricsTest.outputDir),
+      Map(BamMetricsTest.executables.toSeq: _*)) ++
       (if (amplicon) Map("amplicon_bed" -> "amplicon.bed") else Map()) ++
+      (if (rna) Map("transcript_refflat" -> "transcripts.refFlat") else Map()) ++
       Map("regions_of_interest" -> (1 to rois).map("roi_" + _ + ".bed").toList)
     val bammetrics: BamMetrics = initPipeline(map)
 
@@ -59,7 +60,7 @@ class BamMetricsTest extends TestNGSuite with Matchers {
     var regions: Int = rois + (if (amplicon) 1 else 0)
 
     bammetrics.functions.count(_.isInstanceOf[CollectRnaSeqMetrics]) shouldBe (if (rna) 1 else 0)
-    bammetrics.functions.count(_.isInstanceOf[CollectWgsMetrics]) shouldBe 1
+    bammetrics.functions.count(_.isInstanceOf[CollectWgsMetrics]) shouldBe (if (rna) 0 else 1)
     bammetrics.functions.count(_.isInstanceOf[CollectMultipleMetrics]) shouldBe 1
     bammetrics.functions.count(_.isInstanceOf[CalculateHsMetrics]) shouldBe (if (amplicon) 1 else 0)
     bammetrics.functions.count(_.isInstanceOf[CollectTargetedPcrMetrics]) shouldBe (if (amplicon) 1 else 0)

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

@@ -15,6 +15,8 @@ class CollectMultipleMetrics(val root: Configurable) extends Picard with Summari
 
   javaMainClass = new picard.analysis.CollectMultipleMetrics().getClass.getName
 
+  override val defaultCoreMemory = 6.0
+
   @Input(doc = "The input SAM or BAM files to analyze", required = true)
   var input: File = null
 
@@ -41,7 +43,7 @@ class CollectMultipleMetrics(val root: Configurable) extends Picard with Summari
       }
       case _ if p == Programs.CollectInsertSizeMetrics.toString => {
         outputFiles :+= new File(outputName + ".insert_size_metrics")
-        outputFiles :+= new File(outputName + ".insert_size_Histogram.pdf")
+        outputFiles :+= new File(outputName + ".insert_size_histogram.pdf")
       }
       case _ if p == Programs.QualityScoreDistribution.toString => {
         outputFiles :+= new File(outputName + ".quality_distribution_metrics")
@@ -85,17 +87,25 @@ class CollectMultipleMetrics(val root: Configurable) extends Picard with Summari
         case _ => None
       }
       val sum = new Summarizable {
-        override def summaryFiles: Map[String, File] = Map()
         override def summaryStats = stats
+        override def summaryFiles: Map[String, File] = Map()
       }
       qscript.addSummarizable(sum, p)
     })
 
   }
 
-  def summaryFiles = Map()
-
   def summaryStats = Map()
+
+  def summaryFiles = {
+    program.map {
+      case p if p == Programs.CollectInsertSizeMetrics.toString =>
+        Map(
+          "insert_size_histogram" -> new File(outputName + ".insert_size_histogram.pdf"),
+          "insert_size_metrics" -> new File(outputName + ".insert_size_metrics"))
+      case otherwise => Map()
+    }.foldLeft(Map.empty[String, File]) { case (acc, m) => (acc ++ m) }
+  }
 }
 
 object CollectMultipleMetrics {

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

@@ -32,7 +32,7 @@ class CollectRnaSeqMetrics(val root: Configurable) extends Picard with Summariza
   var input: File = null
 
   @Input(doc = "Gene annotations in refFlat form", required = true)
-  var refFlat: File = config("refFlat")
+  var refFlat: File = null
 
   @Input(doc = "Location of rRNA sequences in interval list format", required = false)
   var ribosomalIntervals: Option[File] = config("ribosomal_intervals")
@@ -68,6 +68,7 @@ class CollectRnaSeqMetrics(val root: Configurable) extends Picard with Summariza
   var stopAfter: Option[Long] = config("stop_after")
 
   override def beforeGraph: Unit = {
+    if (refFlat == null) refFlat = config("refFlat")
     val validFlags = StrandSpecificity.values.map(_.toString).toSet
     strandSpecificity match {
       case Some(s) => require(validFlags.contains(s),
@@ -84,7 +85,9 @@ class CollectRnaSeqMetrics(val root: Configurable) extends Picard with Summariza
       "output_chart" -> chartOutput
     ).collect { case (key, Some(value)) => key -> value }
 
-  def summaryStats = Picard.getMetrics(output).getOrElse(Map())
+  def summaryStats = Map(
+    "metrics" -> Picard.getMetrics(output).getOrElse(Map()),
+    "histogram" -> Picard.getHistogram(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/Picard.scala

@@ -82,25 +82,29 @@ abstract class Picard extends BiopetJavaCommandLineFunction {
 object Picard extends Logging {
 
   lazy val getBiopetPicardVersion: Option[String] = {
-    val reader = Source.fromInputStream(getClass.getResourceAsStream("/dependency_list.txt"))
-    val dependencies = reader.getLines().map(_.trim.split(":")).filter(_.size == 5).map(line => Map(
-      "groupId" -> line(0),
-      "artifactId" -> line(1),
-      "type" -> line(2),
-      "version" -> line(3),
-      "scope" -> line(4)
-    )).toList
-
-    logger.debug("dependencies: " + dependencies)
-
-    val htsjdk = dependencies.find(dep => dep("groupId") == "samtools" && dep("artifactId") == "htsjdk").collect {
-      case dep =>
-        "samtools htsjdk " + dep("version")
-    }
+    Option(getClass.getResourceAsStream("/dependency_list.txt")) match {
+      case Some(src) =>
+        val dependencies = Source.fromInputStream(src)
+          .getLines().map(_.trim.split(":")).filter(_.size == 5).map(line => Map(
+            "groupId" -> line(0),
+            "artifactId" -> line(1),
+            "type" -> line(2),
+            "version" -> line(3),
+            "scope" -> line(4)
+          )).toList
+
+        logger.debug("dependencies: " + dependencies)
+
+        val htsjdk = dependencies.find(dep => dep("groupId") == "samtools" && dep("artifactId") == "htsjdk").collect {
+          case dep =>
+            "samtools htsjdk " + dep("version")
+        }
 
-    dependencies.find(dep => dep("groupId") == "picard" && dep("artifactId") == "picard").collect {
-      case dep =>
-        "Picard " + dep("version") + " using " + htsjdk.getOrElse("unknown htsjdk")
+        dependencies.find(dep => dep("groupId") == "picard" && dep("artifactId") == "picard").collect {
+          case dep =>
+            "Picard " + dep("version") + " using " + htsjdk.getOrElse("unknown htsjdk")
+        }
+      case otherwise => None
     }
   }
 

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/tools/BiopetFlagstat.scala

@@ -36,7 +36,7 @@ class BiopetFlagstat(val root: Configurable) extends ToolCommandFuntion with Sum
   @Output(doc = "summary output file", shortName = "output", required = false)
   var summaryFile: File = _
 
-  override val defaultCoreMemory = 2.0
+  override val defaultCoreMemory = 6.0
 
   override def commandLine = super.commandLine + required("-I", input) + required("-s", summaryFile) + " > " + required(output)
 

public/biopet-framework/src/main/scala/nl/lumc/sasc/biopet/tools/MergeTables.scala

@@ -32,7 +32,7 @@ class MergeTables(val root: Configurable) extends ToolCommandFuntion {
 
   javaMainClass = getClass.getName
 
-  override val defaultCoreMemory = 2.0
+  override val defaultCoreMemory = 6.0
 
   /** List of input tabular files */
   @Input(doc = "Input table files", required = true)
@@ -71,7 +71,7 @@ class MergeTables(val root: Configurable) extends ToolCommandFuntion {
       required("-a", valueColumnIndex) +
       optional("-n", idColumnName) +
       optional("-e", fileExtension) +
-      optional("-h", numHeaderLines) +
+      optional("-m", numHeaderLines) +
       optional("-f", fallbackString) +
       optional("-d", delimiter) +
       required("-o", output) +
@@ -164,7 +164,7 @@ object MergeTables extends ToolCommand {
                   idColumnIndices: Seq[Int] = Seq.empty[Int],
                   valueColumnIndex: Int = -1,
                   fileExtension: String = "",
-                  numHeaderLines: Int = 1,
+                  numHeaderLines: Int = 0,
                   fallbackString: String = "-",
                   delimiter: Char = '\t',
                   out: File = new File("-")) extends AbstractArgs
@@ -206,9 +206,9 @@ object MergeTables extends ToolCommand {
       c.copy(fileExtension = x)
     } text "Common extension of all input tables to strip (default: empty string)"
 
-    opt[Int]('h', "num_header_lines") optional () action { (x, c) =>
+    opt[Int]('m', "num_header_lines") optional () action { (x, c) =>
       c.copy(numHeaderLines = x)
-    } text "The number of header lines present in all input files (default: 1; 1-line header)"
+    } text "The number of header lines present in all input files (default: 0; no header)"
 
     opt[String]('f', "fallback") optional () action { (x, c) =>
       c.copy(fallbackString = x)

public/biopet-framework/src/test/scala/nl/lumc/sasc/biopet/tools/MergeTablesTest.scala

@@ -143,7 +143,7 @@ class MergeTablesTest extends TestNGSuite with MockitoSugar with Matchers {
     // default arguments
     parsed.fallbackString shouldBe "-"
     parsed.fileExtension shouldBe ""
-    parsed.numHeaderLines shouldBe 1
+    parsed.numHeaderLines shouldBe 0
     parsed.delimiter shouldBe '\t'
   }
 }

public/gentrap/src/main/resources/nl/lumc/sasc/biopet/pipelines/gentrap/scripts/pdf_report.py

@@ -53,7 +53,7 @@ class FastQCModule(object):
 
     def __repr__(self):
         return '%s(%s)' % (self.__class__.__name__,
-                '[%r, ...]' % self.raw_lines[0])
+                           '[%r, ...]' % self.raw_lines[0])
 
     def __str__(self):
         return ''.join(self.raw_lines)
@@ -88,7 +88,7 @@ class FastQCModule(object):
         self._name = name
         status = tokens[-1]
         assert status in ('pass', 'fail', 'warn'), "Unknown module status: %r" \
-            % status
+                                                   % status
         self._status = status
         # and column names from second line
         columns = self.raw_lines[1][1:].strip().split('\t')
@@ -123,7 +123,7 @@ class FastQC(object):
         '>>Sequence Duplication Levels': 'sequence_duplication_levels',
         '>>Overrepresented sequences': 'overrepresented_sequences',
         '>>Kmer content': 'kmer_content',
-    }
+        }
 
     def __init__(self, fname):
         """
@@ -299,12 +299,12 @@ class LongTable(object):
             "\\hline \\hline",
             "\\endhead",
             "\\hline \\multicolumn{%i}{c}{\\textit{Continued on next page}}\\\\" % \
-                    colnum,
+            colnum,
             "\\hline",
             "\\endfoot",
             "\\hline",
             "\\endlastfoot",
-        ]
+            ]
 
     def __str__(self):
         return "\n".join(self.lines)
@@ -314,7 +314,7 @@ class LongTable(object):
 
     def end(self):
         self.lines.extend(["\\end{longtable}", "\\end{center}",
-            "\\addtocounter{table}{-1}"])
+                           "\\addtocounter{table}{-1}"])
 
 
 # filter functions for the jinja environment
@@ -348,7 +348,7 @@ def float2nice_pct(num, default="None"):
 # and some handy functions
 def natural_sort(inlist):
     key = lambda x: [int(a) if a.isdigit() else a.lower() for a in
-            re.split("([0-9]+)", x)]
+                     re.split("([0-9]+)", x)]
     inlist.sort(key=key)
     return inlist
 
@@ -383,7 +383,7 @@ def write_template(run, template_file, logo_file):
     run.logo = logo_file
     render_vars = {
         "run": run,
-    }
+        }
     rendered = jinja_template.render(**render_vars)
 
     print(rendered, file=sys.stdout)
@@ -417,36 +417,43 @@ class GentrapLib(object):
             self.fastqc_r2_qc_files = self.flexiprep["files"]["fastqc_R2_qc"]
             self.fastqc_r2_qc = FastQC(self.fastqc_r2_qc_files["fastqc_data"]["path"])
         # mapping metrics settings
-        self.aln_metrics = summary.get("bammetrics", {}).get("stats", {}).get("alignment_metrics", {})
+        self.aln_metrics = summary.get("bammetrics", {}).get("stats", {}).get("CollectAlignmentSummaryMetrics", {})
+        for k, v in self.aln_metrics.items():
+            self.aln_metrics[k] = {a.lower(): b for a, b in v.items()}
         # insert size metrics files
-        self.inserts_metrics_files = summary.get("bammetrics", {}).get("files", {}).get("insert_size_metrics", {})
+        self.inserts_metrics_files = \
+            summary.get("bammetrics", {}).get("files", {}).get("multi_metrics", {})
         # rna metrics files and stats
-        self.rna_metrics_files = summary.get("gentrap", {}).get("files", {}).get("rna_metrics", {})
-        _rmetrics = summary.get("gentrap", {}).get("stats", {}).get("rna_metrics", {})
+        self.rna_metrics_files = summary.get("bammetrics", {}).get("files", {}).get("rna", {})
+        _rmetrics = summary.get("bammetrics", {}).get("stats", {}).get("rna", {})
         if _rmetrics:
-            self.rna_metrics = {k: v for k, v in _rmetrics.items() }
+            if "metrics" in _rmetrics:
+                _rmetrics = _rmetrics["metrics"]
+        if _rmetrics:
+            _rmetrics = {k.lower(): v for k, v in _rmetrics.items() }
+            self.rna_metrics = _rmetrics
             pf_bases = float(_rmetrics["pf_bases"])
             exonic_bases = int(_rmetrics.get("coding_bases", 0)) + int(_rmetrics.get("utr_bases", 0))
             # picard uses pct_ but it's actually ratio ~ we follow their convention
             pct_exonic_bases_all = exonic_bases / float(_rmetrics["pf_bases"])
             pct_exonic_bases = exonic_bases / float(_rmetrics.get("pf_aligned_bases", 0))
             self.rna_metrics.update({
-                    "exonic_bases": exonic_bases,
-                    "pct_exonic_bases_all": pct_exonic_bases_all,
-                    "pct_exonic_bases": pct_exonic_bases,
-                    "pct_aligned_bases": 1.0,
-                    "pct_aligned_bases_all": float(_rmetrics.get("pf_aligned_bases", 0.0)) / pf_bases,
-                    "pct_coding_bases_all": float(_rmetrics.get("coding_bases", 0.0)) / pf_bases,
-                    "pct_utr_bases_all": float(_rmetrics.get("utr_bases", 0.0)) / pf_bases,
-                    "pct_intronic_bases_all": float(_rmetrics.get("intronic_bases", 0.0)) / pf_bases,
-                    "pct_intergenic_bases_all": float(_rmetrics.get("intergenic_bases", 0.0)) / pf_bases,
-            })
+                "exonic_bases": exonic_bases,
+                "pct_exonic_bases_all": pct_exonic_bases_all,
+                "pct_exonic_bases": pct_exonic_bases,
+                "pct_aligned_bases": 1.0,
+                "pct_aligned_bases_all": float(_rmetrics.get("pf_aligned_bases", 0.0)) / pf_bases,
+                "pct_coding_bases_all": float(_rmetrics.get("coding_bases", 0.0)) / pf_bases,
+                "pct_utr_bases_all": float(_rmetrics.get("utr_bases", 0.0)) / pf_bases,
+                "pct_intronic_bases_all": float(_rmetrics.get("intronic_bases", 0.0)) / pf_bases,
+                "pct_intergenic_bases_all": float(_rmetrics.get("intergenic_bases", 0.0)) / pf_bases,
+                })
             if _rmetrics.get("ribosomal_bases", "") != "":
                 self.rna_metrics["pct_ribosomal_bases_all"] = float(_rmetrics.get("pf_ribosomal_bases", 0.0)) / pf_bases
 
     def __repr__(self):
         return "{0}(sample=\"{1}\", lib=\"{2}\")".format(
-                self.__class__.__name__, self.sample.name, self.name)
+            self.__class__.__name__, self.sample.name, self.name)
 
 
 class GentrapSample(object):
@@ -458,32 +465,36 @@ class GentrapSample(object):
         self._raw = summary
         self.is_paired_end = summary.get("gentrap", {}).get("stats", {}).get("pipeline", {})["all_paired"]
         # mapping metrics settings
-        self.aln_metrics = summary.get("bammetrics", {}).get("stats", {}).get("alignment_metrics", {})
+        self.aln_metrics = summary.get("bammetrics", {}).get("stats", {}).get("CollectAlignmentSummaryMetrics", {})
+        for k, v in self.aln_metrics.items():
+            self.aln_metrics[k] = {a.lower(): b for a, b in v.items()}
         # insert size metrics files
-        self.inserts_metrics_files = summary.get("bammetrics", {}).get("files", {}).get("insert_size_metrics", {})
+        self.inserts_metrics_files = \
+            summary.get("bammetrics", {}).get("files", {}).get("multi_metrics", {})
         # rna metrics files and stats
-        self.rna_metrics_files = summary.get("gentrap", {}).get("files", {}).get("rna_metrics", {})
-        _rmetrics = summary.get("gentrap", {}).get("stats", {}).get("rna_metrics", {})
+        self.rna_metrics_files = summary.get("bammetrics", {}).get("files", {}).get("rna", {})
+        _rmetrics = summary.get("bammetrics", {}).get("stats", {}).get("rna", {})
         if _rmetrics:
-            self.rna_metrics = {k: v for k, v in _rmetrics.items() }
+            if "metrics" in _rmetrics:
+                _rmetrics = _rmetrics["metrics"]
+        if _rmetrics:
+            _rmetrics = {k.lower(): v for k, v in _rmetrics.items() }
+            self.rna_metrics = _rmetrics
             pf_bases = float(_rmetrics["pf_bases"])
             exonic_bases = int(_rmetrics.get("coding_bases", 0)) + int(_rmetrics.get("utr_bases", 0))
             # picard uses pct_ but it's actually ratio ~ we follow their convention
             pct_exonic_bases_all = exonic_bases / float(_rmetrics["pf_bases"])
             pct_exonic_bases = exonic_bases / float(_rmetrics.get("pf_aligned_bases", 0))
             self.rna_metrics.update({
-                    "exonic_bases": exonic_bases,
-                    "pct_exonic_bases_all": pct_exonic_bases_all,
-                    "pct_exonic_bases": pct_exonic_bases,
-                    "pct_aligned_bases": 1.0,
-                    "pct_aligned_bases_all": float(_rmetrics.get("pf_aligned_bases", 0.0)) / pf_bases,
-                    "pct_coding_bases_all": float(_rmetrics.get("coding_bases", 0.0)) / pf_bases,
-                    "pct_utr_bases_all": float(_rmetrics.get("utr_bases", 0.0)) / pf_bases,
-                    "pct_intronic_bases_all": float(_rmetrics.get("intronic_bases", 0.0)) / pf_bases,
-                    "pct_intergenic_bases_all": float(_rmetrics.get("intergenic_bases", 0.0)) / pf_bases,
-            })
-            if self.run.settings["strand_protocol"] != "non_specific":
-                self.rna_metrics.update({
+                "exonic_bases": exonic_bases,
+                "pct_exonic_bases_all": pct_exonic_bases_all,
+                "pct_exonic_bases": pct_exonic_bases,
+                "pct_aligned_bases": 1.0,
+                "pct_aligned_bases_all": float(_rmetrics.get("pf_aligned_bases", 0.0)) / pf_bases,
+                "pct_coding_bases_all": float(_rmetrics.get("coding_bases", 0.0)) / pf_bases,
+                "pct_utr_bases_all": float(_rmetrics.get("utr_bases", 0.0)) / pf_bases,
+                "pct_intronic_bases_all": float(_rmetrics.get("intronic_bases", 0.0)) / pf_bases,
+                "pct_intergenic_bases_all": float(_rmetrics.get("intergenic_bases", 0.0)) / pf_bases,
                 })
             if _rmetrics.get("ribosomal_bases", "") != "":
                 self.rna_metrics["pct_ribosomal_bases_all"] = float(_rmetrics.get("pf_ribosomal_bases", 0.0)) / pf_bases
@@ -491,7 +502,7 @@ class GentrapSample(object):
         self.lib_names = sorted(summary["libraries"].keys())
         self.libs = \
             {l: GentrapLib(self.run, self, l, summary["libraries"][l]) \
-                for l in self.lib_names}
+             for l in self.lib_names}
 
     def __repr__(self):
         return "{0}(\"{1}\")".format(self.__class__.__name__, self.name)
@@ -521,7 +532,7 @@ class GentrapRun(object):
             ("tophat", "alignment"),
             ("star", "alignment"),
             ("htseqcount", "fragment counting"),
-        ]
+            ]
         self.executables = {}
         for k, desc in executables:
             in_summary = self.all_executables.get(k)
@@ -543,7 +554,7 @@ class GentrapRun(object):
         self.sample_names = sorted(summary["samples"].keys())
         self.samples = \
             {s: GentrapSample(self, s, summary["samples"][s]) \
-                for s in self.sample_names}
+             for s in self.sample_names}
         self.libs = []
         for sample in self.samples.values():
             self.libs.extend(sample.libs.values())
@@ -556,19 +567,20 @@ class GentrapRun(object):
 
     def __repr__(self):
         return "{0}(\"{1}\")".format(self.__class__.__name__,
-                                        self.summary_file)
+                                     self.summary_file)
 
 
 if __name__ == "__main__":
 
     parser = argparse.ArgumentParser()
     parser.add_argument("summary_file", type=str,
-            help="Path to Gentrap summary file")
+                        help="Path to Gentrap summary file")
     parser.add_argument("template_file", type=str,
-            help="Path to main template file")