Use interval start as position for Ensembl VEP ANN attributes.

bigdatagenomics · Oct 26, 2020 · 2d9f1c8 · 2d9f1c8
1 parent e63137b
commit 2d9f1c8
Show file tree

Hide file tree

Showing 2 changed files with 25 additions and 2 deletions.
diff --git a/adam-core/src/main/scala/org/bdgenomics/adam/converters/TranscriptEffectConverter.scala b/adam-core/src/main/scala/org/bdgenomics/adam/converters/TranscriptEffectConverter.scala
@@ -83,6 +83,18 @@ private[adam] object TranscriptEffectConverter extends Serializable with Logging
     s.split("&").map(MESSAGES.get(_)).toList.flatten
   }
 
+  /**
+   * Ensembl VEP incorrectly supplies an interval instead of a position
+   * for some attributes; in these cases use the interval start as the position.
+   *
+   * @param s position or interval
+   * @return position or interval start
+   */
+  private def positionOrIntervalStart(s: String): Int = {
+    val tokens = s.split("-")
+    Integer.parseInt(tokens(0))
+  }
+
   /**
    * Split a single or fractional value into optional numerator and denominator values.
    *
@@ -96,8 +108,8 @@ private[adam] object TranscriptEffectConverter extends Serializable with Logging
     val tokens = s.split("/")
     tokens.length match {
       case 0 => (None, None)
-      case 1 => (Some(Integer.parseInt(tokens(0))), None)
-      case _ => (Some(Integer.parseInt(tokens(0))), Some(Integer.parseInt(tokens(1))))
+      case 1 => (Some(positionOrIntervalStart(tokens(0))), None)
+      case _ => (Some(positionOrIntervalStart(tokens(0))), Some(Integer.parseInt(tokens(1))))
     }
   }
 

diff --git a/adam-core/src/test/scala/org/bdgenomics/adam/converters/TranscriptEffectConverterSuite.scala b/adam-core/src/test/scala/org/bdgenomics/adam/converters/TranscriptEffectConverterSuite.scala
@@ -35,6 +35,7 @@ class TranscriptEffectConverterSuite extends ADAMFunSuite {
   final val INVALID_NUMBER = "T|upstream_gene_variant||TAS1R3|ENSG00000169962|transcript|ENST00000339381.5|protein_coding|1/2|c.-485C>T|||4|1/42|not a number|"
   final val INVALID_FRACTION = "T|upstream_gene_variant||TAS1R3|ENSG00000169962|transcript|ENST00000339381.5|protein_coding|not a number/2|c.-485C>T|||4|1/42|453|"
   final val VALID = "T|upstream_gene_variant||TAS1R3|ENSG00000169962|transcript|ENST00000339381.5|protein_coding|1/2|c.-485C>T|||4|1/42|453|"
+  final val VEP_POSITION = "T|upstream_gene_variant||TAS1R3|ENSG00000169962|transcript|ENST00000339381.5|protein_coding|1/2|c.-485C>T|||4-5/420|1/42|453|"
   final val DIFFERENT_ALT = "A|upstream_gene_variant||TAS1R3|ENSG00000169962|transcript|ENST00000339381.5|protein_coding|1/2|c.-485C>T|||4|1/42|453|"
 
   var variant: Variant = null
@@ -170,6 +171,16 @@ class TranscriptEffectConverterSuite extends ADAMFunSuite {
     })
   }
 
+  test("parse VCF ANN attribute with Ensembl VEP position attribute") {
+    val ann = TranscriptEffectConverter.parseAnn(Seq(VEP_POSITION), ValidationStringency.STRICT)
+    assert(ann.length == 1)
+
+    val te = ann.head
+    assert(te.getAlternateAllele == "T")
+    assert(te.getCdsPosition == 4)
+    assert(te.getCdsLength == 420)
+  }
+
   test("convert to transcript effect from null VCF ANN attribute in variant context") {
     when(variantContext.getAttributeAsList("ANN")).thenReturn(null)