Merge 81ed51f into 988897e

bigdatagenomics · Jun 12, 2019 · c2caab7 · c2caab7
2 parents 988897e + 81ed51f
commit c2caab7
Show file tree

Hide file tree

Showing 3 changed files with 24 additions and 7 deletions.
diff --git a/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformFeatures.scala b/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformFeatures.scala
@@ -35,12 +35,16 @@ object TransformFeatures extends BDGCommandCompanion {
 class TransformFeaturesArgs extends Args4jBase with ParquetSaveArgs {
   @Argument(required = true, metaVar = "INPUT",
     usage = "The feature file to convert (e.g., .bed, .gff/.gtf, .gff3, .interval_list, .narrowPeak). If extension is not detected, Parquet is assumed.", index = 0)
-  var featuresFile: String = _
+  var featuresPath: String = _
 
   @Argument(required = true, metaVar = "OUTPUT",
     usage = "Location to write ADAM feature data. If extension is not detected, Parquet is assumed.", index = 1)
   var outputPath: String = null
 
+  @Args4jOption(required = false, name = "-reference",
+    usage = "Load reference for features; .dict as HTSJDK sequence dictionary format, .genome as Bedtools genome file format, .txt as UCSC Genome Browser chromInfo files.")
+  var referencePath: String = null
+
   @Args4jOption(required = false, name = "-num_partitions",
     usage = "Number of partitions to load a text file using.")
   var numPartitions: Int = _
@@ -62,10 +66,12 @@ class TransformFeatures(val args: TransformFeaturesArgs)
   def run(sc: SparkContext) {
     checkWriteablePath(args.outputPath, sc.hadoopConfiguration)
 
+    val optSequenceDictionary = Option(args.referencePath).map(sc.loadSequenceDictionary(_))
+
     sc.loadFeatures(
-      args.featuresFile,
-      optMinPartitions = Option(args.numPartitions),
-      optProjection = None
+      args.featuresPath,
+      optSequenceDictionary = optSequenceDictionary,
+      optMinPartitions = Option(args.numPartitions)
     ).save(args.outputPath, args.single, args.disableFastConcat)
   }
 }
diff --git a/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformSequences.scala b/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformSequences.scala
@@ -41,6 +41,10 @@ class TransformSequencesArgs extends Args4jBase with ParquetSaveArgs {
     usage = "Location to write ADAM sequence data. If extension is not detected, Parquet is assumed.", index = 1)
   var outputPath: String = null
 
+  @Args4jOption(required = false, name = "-create_reference",
+    usage = "Create reference from sequence names and lengths. Defaults to false.")
+  var createReference: Boolean = false
+
   @Args4jOption(required = false, name = "-single",
     usage = "Save as a single file, for the text formats.")
   var single: Boolean = false
@@ -66,6 +70,7 @@ class TransformSequences(val args: TransformSequencesArgs)
       case Alphabet.PROTEIN => sc.loadProteinSequences(args.sequencesFile, optPredicate = None, optProjection = None)
       case Alphabet.RNA     => sc.loadRnaSequences(args.sequencesFile, optPredicate = None, optProjection = None)
     }
-    sequences.save(args.outputPath, args.single, args.disableFastConcat)
+    val maybeCreateReference = if (args.createReference) sequences.createSequenceDictionary() else sequences
+    maybeCreateReference.save(args.outputPath, args.single, args.disableFastConcat)
   }
 }
diff --git a/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformSlices.scala b/adam-cli/src/main/scala/org/bdgenomics/adam/cli/TransformSlices.scala
@@ -44,6 +44,10 @@ class TransformSlicesArgs extends Args4jBase with ParquetSaveArgs {
     usage = "Maximum slice length. Defaults to 10000L.")
   var maximumLength: Long = 10000L
 
+  @Args4jOption(required = false, name = "-create_reference",
+    usage = "Create reference from sequence names and lengths. Defaults to false.")
+  var createReference: Boolean = false
+
   @Args4jOption(required = false, name = "-single",
     usage = "Save as a single file, for the text formats.")
   var single: Boolean = false
@@ -59,11 +63,13 @@ class TransformSlices(val args: TransformSlicesArgs)
   val companion = TransformSlices
 
   def run(sc: SparkContext) {
-    sc.loadSlices(
+    val slices = sc.loadSlices(
       args.slicesFile,
       maximumLength = args.maximumLength,
       optPredicate = None,
       optProjection = None
-    ).save(args.outputPath, args.single, args.disableFastConcat)
+    )
+    val maybeCreateReference = if (args.createReference) slices.createSequenceDictionary() else slices
+    maybeCreateReference.save(args.outputPath, args.single, args.disableFastConcat)
   }
 }