Merge pull request #204 from monarch-initiative/large_variant

Merging after seeing passing tests. I must force-merge because tests for 3.12 will never run on this PR but they are required to meet the specs.

src/genophenocorr/io.py

@@ -87,7 +87,8 @@ def default(self, o):
                 'label': o.label,
                 'meta_label': o.meta_label,
             }
-        elif isinstance(o, (Genotype, VariantEffect, Strand)):
+        elif isinstance(o, (Genotype, VariantEffect, Strand, VariantClass)):
+            # enums
             return o.name
         elif isinstance(o, Phenotype):
             return {
@@ -220,7 +221,7 @@ def object_hook(obj: typing.Dict[typing.Any, typing.Any]) -> typing.Any:
         elif GenophenocorrJSONDecoder._has_all_fields(obj, _IMPRECISE_SV_INFO_FIELDS):
             return ImpreciseSvInfo(
                 structural_type=hpotk.TermId.from_curie(obj['structural_type']),
-                variant_class=obj['variant_class'],
+                variant_class=VariantClass[obj['variant_class']],
                 gene_id=obj['gene_id'],
                 gene_symbol=obj['gene_symbol'],
             )

src/genophenocorr/model/_variant.py

@@ -461,6 +461,8 @@ def variant_key(self) -> str:
         Get a readable representation of the variant's coordinates.
 
         For instance, ``X_12345_12345_C_G`` for a sequence variant or ``22_10001_20000_INV`` for a symbolic variant.
+        If the key is larger than 50 characters, the 'ref' and/or 'alt' (if over 10 bps) are changed to just show number of bps. 
+        Example: ``X_1000001_1000027_TAAAAAAAAAAAAAAAAAAAAAAAAAA_T`` -> ``X_1000001_1000027_--27bp--_T``
 
         .. note::
 
@@ -471,13 +473,15 @@ def variant_key(self) -> str:
         else:
             key = f'{self.chrom}_{self.start + 1}_{self.end}_{self.ref}_{self.alt}'
             if len(key) > 50:
-                ref = None
-                alt = None
                 if len(self.ref) > 10:
                     ref = f"--{len(self.ref)}bp--"
+                else:
+                    ref = self.ref
                 if len(self.alt) > 10:
                     alt = f"--{len(self.alt)}bp--"
-                return f"{self.chrom}_{self.start + 1}_{self.end}_{ref if not None else self.ref}_{alt if not None else self.alt}"
+                else:
+                    alt = self.alt
+                return f"{self.chrom}_{self.start + 1}_{self.end}_{ref}_{alt}"
             else:
                 return key
 

src/genophenocorr/preprocessing/_generic.py

@@ -32,6 +32,7 @@ def annotate(self, item: ImpreciseSvInfo) -> typing.Sequence[TranscriptAnnotatio
                 variant_effects=variant_effects,
                 affected_exons=affected_exons,
                 protein_id=None,
+                hgvsp=None,
                 protein_effect_coordinates=None,
             )
             tx_annotations.append(annotation)

tests/analysis/test_examples.py

@@ -65,8 +65,8 @@ def test_get_count(
         all_counts = results.all_counts
         assert isinstance(all_counts, typing.Mapping)
 
-        # We tested 69 HPO terms
-        assert results.total_tests == 69
+        # We tested 74 HPO terms
+        assert len(all_counts) == 74
 
         # The index of all_counts is a Tuple with (HPO TermId, BooleanPredicate
         # Let's test Seizure - we should have one row for each Patient Predicate

tests/model/test_cohort.py

@@ -16,6 +16,8 @@ def test_all_transcript_ids(
             "NM_001032387.2",
             "NM_001351089.2",
             "NM_000456.3",
+            "NM_016373.4",
+            "NM_013275.6"
         }
 
     def test_variant_effect_count_by_tx(

tests/model/test_variant.py

@@ -2,7 +2,9 @@
 
 import typing
 
-from genophenocorr.model import Variant, Cohort
+from genophenocorr.model import Variant, Cohort, VariantCoordinates
+from genophenocorr.model.genome import GenomeBuild, GenomicRegion, Strand
+
 
 class TestVariant:
 
@@ -27,4 +29,38 @@ def test_get_hgvs_cdna_by_tx(
     ):
         hgvs = some_variant.get_hgvs_cdna_by_tx_id(transcript_id=tx_id)
 
-        assert hgvs == expected
+        assert hgvs == expected
+
+
+class TestVariantCoordinates:
+
+    @pytest.mark.parametrize(
+        "contig_name, start, end, ref, alt, change_length, expected",
+        [
+            ("chrX", 100, 101, "C", "T", 0, "X_101_101_C_T"),
+            ("chrY", 150, 152, "G", "GG", 1, "Y_151_152_G_GG"),
+            ("chr16", 1000, 1040, "A", "GGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGG", 39, "16_1001_1040_A_--40bp--"),
+            ("chr2", 200, 301, "N", "<DEL>", 100, "2_201_301_DEL")
+        ]
+    )
+    def test_variant_key(
+        self,
+        genome_build: GenomeBuild,
+        contig_name: str,
+        start: int, end: int,
+        ref: str, alt: str,
+        change_length: int,
+        expected: str,
+    ):
+        contig = genome_build.contig_by_name(contig_name)
+        assert contig is not None
+
+        vc = VariantCoordinates(
+            region=GenomicRegion(
+                contig=contig,
+                start=start, end=end, strand=Strand.POSITIVE,
+            ),
+            ref=ref, alt=alt, change_length=change_length,
+        )
+
+        assert vc.variant_key == expected

tests/view/test_formatter.py

@@ -11,6 +11,7 @@ class TestFormatter:
         "variant, expected",
         [
             ('12_56004525_56004525_A_G', "NM_001032386.2:c.1136A>G"),
+            ('16_78386858_78425054_--38197bp--_A', 'NM_016373.4:c.517_791del')
         ]
     )
     def test_variant_formatter(