fix(clip): never auto-clip RG, MI and other non-per-base tags

src/main/scala/com/fulcrumgenomics/bam/ClipBam.scala

@@ -82,7 +82,7 @@ class ClipBam
   @arg(flag='m', doc="Optional output of clipping metrics.") val metrics: Option[FilePath] = None,
   @arg(flag='r', doc="Reference sequence fasta file.") val ref: PathToFasta,
   @arg(flag='c', doc="The type of clipping to perform.") val clippingMode: ClippingMode = ClippingMode.Hard,
-  @arg(flag='a', doc="Automatically clip extended attributes that are the same length as bases.") val autoClipAttributes: Boolean = false,
+  @arg(flag='a', doc="Automatically clip extended attributes that are the same length as bases, except listed tags that are not per-base data for the read, such as RG.") val autoClipAttributes: Boolean = false,
   @arg(flag='H', doc="Upgrade all existing clipping in the input to the given clipping mode prior to applying any other clipping criteria.") val upgradeClipping: Boolean = false,
   @arg(          doc="Require at least this number of bases to be clipped on the 5' end of R1") val readOneFivePrime: Int  = 0,
   @arg(          doc="Require at least this number of bases to be clipped on the 3' end of R1") val readOneThreePrime: Int = 0,

src/main/scala/com/fulcrumgenomics/bam/SamRecordClipper.scala

@@ -54,6 +54,17 @@ object SamRecordClipper {
   /** The set of tags that should be invalidated if a read undergoes clipping. */
   val TagsToInvalidate: Seq[String] = Bams.AlignmentTags
 
+  /** Tags whose values are not this read's per-base data, so they are not auto-clipped even when their length equals
+    * the read's. Best-effort: an unlisted tag whose length matches the read's is still clipped. */
+  val TagsNeverAutoClipped: Set[String] = Set(
+    "RG", "LB", "PU", "PG", "CO", "MI",
+    "BC", "QT", "RX", "QX", "OX", "BZ", "CB", "CR", "CY", "UB", "UR", "UY", "BX",
+    "MC", "MD", "SA", "OA", "OC", "CG", "XA", "cs", "jM", "jI",
+    "CC", "CT", "FS", "PT", "GX", "GN",
+    "R2", "Q2",
+    "FZ", "MM", "ML", "mv", "pi", "st", "fn"
+  )
+
   private val NoCallBase = 'N'.toByte
   private val NoCallQual = 2.toByte
 }
@@ -76,7 +87,8 @@ object SamRecordClipper {
   * @param autoClipAttributes if true attributes that are the same length as the bases
   *                           and qualities be automatically clipped in the same way as
   *                           the bases and qualities, otherwise attributes are not
-  *                           touched.
+  *                           touched. Tags in [[SamRecordClipper.TagsNeverAutoClipped]]
+  *                           are never clipped.
   */
 class SamRecordClipper(val mode: ClippingMode, val autoClipAttributes: Boolean) {
   import SamRecordClipper._
@@ -491,6 +503,7 @@ class SamRecordClipper(val mode: ClippingMode, val autoClipAttributes: Boolean)
       val newLength = rec.length
       val oldLength = newLength + remove
       rec.attributes.foreach {
+        case (tag, _) if TagsNeverAutoClipped.contains(tag) => ()
         case (tag, s: String)   if s.length == oldLength => rec(tag) = if (fromStart) s.drop(remove) else s.take(newLength)
         case (tag, a: Array[_]) if a.length == oldLength => rec(tag) = if (fromStart) a.drop(remove) else a.take(newLength)
         case _  => ()

src/main/scala/com/fulcrumgenomics/bam/TrimPrimers.scala

@@ -82,7 +82,7 @@ class TrimPrimers
   @arg(flag='S', doc="Match to primer locations +/- this many bases.") val slop: Int = 5,
   @arg(flag='s', doc="Sort order of output BAM file (defaults to input sort order).") val sortOrder: Option[SamOrder] = None,
   @arg(flag='r', doc="Optional reference fasta for recalculating NM, MD and UQ tags.") val ref: Option[PathToFasta] = None,
-  @arg(flag='a', doc="Automatically trim extended attributes that are the same length as bases.") val autoTrimAttributes: Boolean = false,
+  @arg(flag='a', doc="Automatically trim extended attributes that are the same length as bases, except listed tags that are not per-base data for the read, such as RG.") val autoTrimAttributes: Boolean = false,
   @arg(doc="Trim only first of pair reads (R1s) or fragment reads, otherwise both ends of a pair.") val firstOfPair: Boolean = false
 
 )extends FgBioTool with LazyLogging {

src/test/scala/com/fulcrumgenomics/bam/SamRecordClipperTest.scala

@@ -535,6 +535,31 @@ class SamRecordClipperTest extends UnitSpec with OptionValues {
     withAuto[String]("az")  shouldBe "67890123456789012345678901234567890"
   }
 
+  it should "not auto-clip tags that are not this read's per-base data, even when their length equals the read's" in {
+    val expected = Set(
+      "RG", "LB", "PU", "PG", "CO", "MI",
+      "BC", "QT", "RX", "QX", "OX", "BZ", "CB", "CR", "CY", "UB", "UR", "UY", "BX",
+      "MC", "MD", "SA", "OA", "OC", "CG", "XA", "cs", "jM", "jI",
+      "CC", "CT", "FS", "PT", "GX", "GN",
+      "R2", "Q2",
+      "FZ", "MM", "ML", "mv", "pi", "st", "fn"
+    )
+    SamRecordClipper.TagsNeverAutoClipped shouldBe expected
+
+    val rec     = r(10, "5S10M5S")
+    val value   = "ACGTACGTACGTACGTACGT"
+    val depth   = Array.tabulate[Short](20)(_.toShort)
+    val perBase = Seq("OQ", "E2", "XB")
+    expected.foreach(tag => rec(tag) = value)
+    perBase.foreach(tag => rec(tag) = value)
+    rec("cd") = depth
+
+    clipper(Hard, autoClip=true).upgradeAllClipping(rec) shouldBe (5, 5)
+    expected.foreach(tag => rec[String](tag) shouldBe value)
+    perBase.foreach(tag => rec[String](tag) shouldBe value.slice(5, 15))
+    rec[Array[Short]]("cd") shouldBe depth.slice(5, 15)
+  }
+
   it should "not convert reads that have no soft-clipping" in {
     val noSoft = r(10, "55M", attrs=Map("az" -> "12345678901234567890123456789012345678901234567890"))
     val hard   = r(10, "5H55M10H", attrs=Map("az" -> "12345678901234567890123456789012345678901234567890"))