Merge pull request #52 from NCI-GDC/feat/variant-support

added a VariantSupport column type for RNA support maf column

maflib/column_types.py

@@ -26,6 +26,7 @@
     StrandEnum,
     ValidationStatusEnum,
     VariantClassificationEnum,
+    VariantSupportEnum,
     VariantTypeEnum,
     VerificationStatusEnum,
     YesNoOrUnknownEnum,
@@ -552,6 +553,14 @@ def __enum_class__(cls) -> Type[VariantTypeEnum]:
         return VariantTypeEnum
 
 
+class VariantSupport(EnumColumn):
+    """A column that represents the RNA_Support MAF column"""
+
+    @classmethod
+    def __enum_class__(cls) -> Type[VariantSupportEnum]:
+        return VariantSupportEnum
+
+
 class VerificationStatus(NullableEmptyStringIsNone, EnumColumn):
     """A column that represents the 'Verification_Status' MAF column,
     where the empty string is treated as null."""

maflib/column_values.py

@@ -95,6 +95,16 @@ class VariantTypeEnum(MafEnum):
     Consolidated: str = "Consolidated"
 
 
+@unique
+class VariantSupportEnum(MafEnum):
+    """Enumeration for "Match", "Overlap", "No", "Unknown" column values."""
+
+    Match = "Match"
+    Overlap = "Overlap"
+    No = "No"
+    Unknown = "Unknown"
+
+
 @unique
 class VerificationStatusEnum(MafEnum):
     """Enumeration for the MAF 'Verification_Status' column value"""

maflib/schemas/gdc-2.0.0-aliquot-merged.json

@@ -3,12 +3,12 @@
 	"annotation-spec" : "gdc-2.0.0-aliquot-merged",
 	"extends" : "gdc-2.0.0-aliquot",
 	"columns" : [
-            [ "RNA_Support", "YesNoOrUnknown", "Indicates if the variant is supported (Yes) or not supported (No) by tumor RNA-Seq. If it has not been checked against RNA-Seq data, the value will be 'Unknown'." ],
+            [ "RNA_Support", "VariantSupport", "Indicates if the variant is found and alleles (Match), simply (Overlap), or is not supported (No) by tumor RNA-Seq. If it has not been checked against RNA-Seq data, the value will be 'Unknown'." ],
             [ "RNA_depth", "NullableZeroBasedIntegerColumn", "Read depth at this locus if the variant is supported by tumor RNA-seq data." ],
             [ "RNA_ref_count", "NullableZeroBasedIntegerColumn", "Read depth supporting the reference allele at this locus if the variant is supported by tumor RNA-seq data." ],
             [ "RNA_alt_count", "NullableZeroBasedIntegerColumn", "Read depth supporting the variant allele at this locus if the variant is supported by tumor RNA-seq data." ],
             [ "callers", "SequenceOfStrings", "A semicolon delimited list of supporting variant callers. If an asterisk is added to the caller name, this indicates the locus is a complex variant and this caller overlapped but did not support (e.g., not all callers can detect MNPs)" ]
 	],
 	"filtered" : ["ALLELE_NUM", "MINIMISED", "FILTER", "src_vcf_id", "vcf_region",
-                  "vcf_info", "vcf_format", "vcf_tumor_gt", "vcf_normal_gt"] 
+                  "vcf_info", "vcf_format", "vcf_tumor_gt", "vcf_normal_gt"]
 }

tests/maflib/test_column_types.py

@@ -487,6 +487,33 @@ def test_invalid(self):
             column_types.YesNoOrUnknown.build("key", "Who Knows")
 
 
+class TestVariantSupport(TestCase):
+    def test_valid(self):
+        self.is_column_is_valid(
+            column_types.VariantSupport.build("key", "Match"),
+            column_values.VariantSupportEnum.Match,
+        )
+        self.is_column_is_valid(
+            column_types.VariantSupport.build("key", "Overlap"),
+            column_values.VariantSupportEnum.Overlap,
+        )
+        self.is_column_is_valid(
+            column_types.VariantSupport.build("key", "No"),
+            column_values.VariantSupportEnum.No,
+        )
+        self.is_column_is_valid(
+            column_types.VariantSupport.build("key", "Unknown"),
+            column_values.VariantSupportEnum.Unknown,
+        )
+
+    def test_invalid(self):
+        # "Foo" is an invalid value so ValueError is raised but is caught and
+        # the fallback attempts to use "Foo" to lookup the correct Enum value
+        # hence generating a KeyError
+        with self.assertRaises(KeyError):
+            column_types.VariantSupport.build("key", "Foo")
+
+
 class TestNullableYOrN(TestCase):
     def test_valid(self):
         nulls = [
@@ -594,6 +621,7 @@ class TestCustomEnums(TestCase):
         column_types.Impact,
         column_types.MC3Overlap,
         column_types.GdcValidationStatus,
+        column_types.VariantSupport,
     ]
 
     def test_valid(self):