[ADAM-351] Add simple support for symbolic alleles.

bigdatagenomics · Aug 16, 2014 · 5afa75b · 5afa75b
1 parent 49eac57
commit 5afa75b
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 8 deletions.
diff --git a/adam-core/src/main/scala/org/bdgenomics/adam/converters/VariantContextConverter.scala b/adam-core/src/main/scala/org/bdgenomics/adam/converters/VariantContextConverter.scala
@@ -97,27 +97,29 @@ class VariantContextConverter(dict: Option[SequenceDictionary] = None) extends S
         return Seq(ADAMVariantContext(variant, genotypes, None))
       }
       case List(allele) => {
-        assert(allele.isNonReference && !allele.isSymbolic)
-        val variant = createADAMVariant(vc, Some(allele.getBaseString))
-        val genotypes = extractNonReferenceGenotypes(vc, variant, calling_annotations)
+        assert(allele.isNonReference,
+          "Assertion failed when converting: " + vc.toString)
+        val variant = createADAMVariant(vc, Some(allele.getDisplayString))
+        val genotypes = extractReferenceModelGenotypes(vc, variant, calling_annotations)
         return Seq(ADAMVariantContext(variant, genotypes, None))
       }
       case List(allele, NON_REF_ALLELE) => {
-        assert(allele.isNonReference && !allele.isSymbolic)
-        val variant = createADAMVariant(vc, Some(allele.getBaseString))
+        assert(allele.isNonReference,
+          "Assertion failed when converting: " + vc.toString)
+        val variant = createADAMVariant(vc, Some(allele.getDisplayString))
         val genotypes = extractReferenceModelGenotypes(vc, variant, calling_annotations)
         return Seq(ADAMVariantContext(variant, genotypes, None))
       }
       case alleles :+ NON_REF_ALLELE => {
-        assert(false, "Multi-allelic site with non-ref symbolic allele")
-        Seq()
+        throw new IllegalArgumentException("Multi-allelic site with non-ref symbolic allele" +
+          vc.toString)
       }
       case _ => {
         // Default case is multi-allelic without reference model
         val vcb = new VariantContextBuilder(vc)
         return vc.getAlternateAlleles.flatMap(allele => {
           val idx = vc.getAlleleIndex(allele)
-          assert(idx >= 1, "Unexpected index for alternate allele")
+          assert(idx >= 1, "Unexpected index for alternate allele: " + vc.toString)
           vcb.alleles(List(vc.getReference, allele, NON_REF_ALLELE))
 
           def punchOutGenotype(g: org.broadinstitute.variant.variantcontext.Genotype, idx: Int): org.broadinstitute.variant.variantcontext.Genotype = {

diff --git a/adam-core/src/test/scala/org/bdgenomics/adam/converters/VariantContextConverterSuite.scala b/adam-core/src/test/scala/org/bdgenomics/adam/converters/VariantContextConverterSuite.scala
@@ -49,6 +49,12 @@ class VariantContextConverterSuite extends FunSuite {
     .stop(1L)
     .chr("1")
 
+  def gatkCNVBuilder: VariantContextBuilder = new VariantContextBuilder()
+    .alleles(List(Allele.create("A", true), Allele.create("<CN0>", false)))
+    .start(10L)
+    .stop(20L)
+    .chr("1")
+
   def adamSNVBuilder(contig: String = "1"): Variant.Builder = Variant.newBuilder()
     .setContig(Contig.newBuilder().setContigName(contig).build())
     .setStart(0L)
@@ -82,6 +88,24 @@ class VariantContextConverterSuite extends FunSuite {
     assert(variant.getContig.getContigName === "NC_000001.10")
   }
 
+  test("Convert GATK site-only CNV to ADAM") {
+    val converter = new VariantContextConverter
+
+    val adamVCs = converter.convert(gatkCNVBuilder.make)
+    assert(adamVCs.length === 1)
+    val adamVC = adamVCs.head
+
+    assert(adamVC.genotypes.size === 0)
+
+    val variant = adamVC.variant
+    assert(variant.getContig.getContigName === "1")
+
+    assert(variant.getReferenceAllele === "A")
+    assert(variant.getAlternateAllele === "<CN0>")
+    assert(variant.getStart === 9L)
+    assert(variant.getEnd === 20L)
+  }
+
   test("Convert GATK SNV w/ genotypes w/ phase information to ADAM") {
     val vcb = gatkSNVBuilder