Stores original position and original cigar during realignment.

adam-core/src/main/scala/org/bdgenomics/adam/converters/AlignmentRecordConverter.scala

@@ -115,7 +115,8 @@ class AlignmentRecordConverter extends Serializable {
 
  // set the cigar, if provided
  Option(adamRecord.getCigar).map(_.toString).foreach(builder.setCigarString)
-
+ // set the old cigar, if provided
+ Option(adamRecord.getOldCigar).map(_.toString).foreach(v => builder.setAttribute("OC", v))
  // set mapping flags
  Option(adamRecord.getReadNegativeStrand)
  .foreach(v => builder.setReadNegativeStrandFlag(v.booleanValue))
@@ -125,6 +126,8 @@ class AlignmentRecordConverter extends Serializable {
  .foreach(v => builder.setSupplementaryAlignmentFlag(v.booleanValue))
  Option(adamRecord.getStart)
  .foreach(s => builder.setAlignmentStart(s.toInt + 1))
+ Option(adamRecord.getOldPosition)
+ .foreach(s => builder.setAttribute("OP", s.toInt + 1))
  Option(adamRecord.getMapq).foreach(v => builder.setMappingQuality(v))
  } else {
  // mapping quality must be 0 if read is unmapped

adam-core/src/main/scala/org/bdgenomics/adam/converters/SAMRecordConverter.scala

@@ -66,6 +66,12 @@ class SAMRecordConverter extends Serializable {
  assert(start != 0, "Start cannot equal 0 if contig is set.")
  builder.setStart((start - 1).asInstanceOf[Long])
 
+ // set OP and OC flags, if applicable
+ if (samRecord.getAttribute("OP") != null) {
+ builder.setOldPosition(samRecord.getIntegerAttribute("OP").toLong - 1)
+ builder.setOldCigar(samRecord.getStringAttribute("OC"))
+ }
+
  val end = samRecord.getCigar.getCigarElements
  .asScala
  .filter((p: CigarElement) => p.getOperator.consumesReferenceBases())

adam-core/src/main/scala/org/bdgenomics/adam/rdd/read/realignment/RealignIndels.scala

@@ -357,6 +357,8 @@ private[rdd] class RealignIndels(val consensusModel: ConsensusGenerator = new Co
 
  // update mdtag and cigar
  builder.setMismatchingPositions(MdTag.moveAlignment(r, newCigar, reference.drop(remapping), refStart + remapping).toString())
+ builder.setOldPosition(r.getStart())
+ builder.setOldCigar(r.getCigar())
  builder.setCigar(newCigar.toString)
  new RichAlignmentRecord(builder.build())
  } else

adam-core/src/test/scala/org/bdgenomics/adam/converters/AlignmentRecordConverterSuite.scala

@@ -43,6 +43,8 @@ class AlignmentRecordConverterSuite extends FunSuite {
  .setMapq(60)
  .setQual(sequence) // no typo, we just don't care
  .setMismatchingPositions(mdtag)
+ .setOldPosition(12)
+ .setOldCigar("2^AAA3")
  .build()
  }
 
@@ -82,6 +84,8 @@ class AlignmentRecordConverterSuite extends FunSuite {
  assert(toSAM.getMappingQuality === 60)
  assert(toSAM.getBaseQualityString === sequence)
  assert(toSAM.getAttribute("MD") === "2^AAA2")
+ assert(toSAM.getIntegerAttribute("OP") === 13)
+ assert(toSAM.getStringAttribute("OC") === "2^AAA3")
  }
 
  test("convert a read to fastq") {

adam-core/src/test/scala/org/bdgenomics/adam/rdd/read/realignment/RealignIndelsSuite.scala

@@ -254,4 +254,18 @@ class RealignIndelsSuite extends SparkFunSuite {
  // this should be a NOP
  assert(RealignIndels(reads).count === 4)
  }
+
+ sparkTest("test OP and OC tags") {
+ artificial_realigned_reads.collect().foreach(realn => {
+ val readName = realn.getReadName()
+ val op = realn.getOldPosition()
+ val oc = realn.getOldCigar()
+
+ Option(op).filter(_ >= 0).foreach(oPos => {
+ val s = artificial_reads.collect().filter(x => (x.getReadName() == readName))
+ assert(s.filter(x => (x.getStart() == oPos)).length > 0)
+ assert(s.filter(x => (x.getCigar() == oc)).length > 0)
+ })
+ })
+ }
 }

pom.xml

@@ -305,7 +305,7 @@
  <dependency>
  <groupId>org.bdgenomics.bdg-formats</groupId>
  <artifactId>bdg-formats</artifactId>
- <version>0.3.1</version>
+ <version>0.3.2</version>
  </dependency>
  <dependency>
  <groupId>commons-io</groupId>