varcode 9.2.0__tar.gz → 9.2.1__tar.gz

{varcode-9.2.0/varcode.egg-info → varcode-9.2.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: varcode
-Version: 9.2.0
+Version: 9.2.1
 Summary: Variant annotation in Python
 Author-email: Alex Rubinsteyn <alex.rubinsteyn@unc.edu>
 Project-URL: Homepage, https://github.com/openvax/varcode

varcode-9.2.1/tests/test_symbolic_sv_padding.py

@@ -0,0 +1,81 @@
+"""Symbolic POS is retained; only POS+1..END changes (#404, VCF 4.3)."""
+
+from types import SimpleNamespace
+
+import pytest
+from pyensembl import cached_release
+
+from varcode import StructuralVariant, parse_symbolic_alt, load_vcf
+from varcode.effects.structural import (
+    _build_deletion_mutant_transcript, _build_duplication_mutant_transcript,
+    _build_inversion_mutant_transcript,
+)
+from varcode.nucleotides import reverse_complement
+
+
+@pytest.mark.parametrize("kind", ["DEL", "DUP", "INV", "CNV", "CN0", "CN3"])
+def test_parsed_span_excludes_padding_but_keeps_record_coordinates(kind):
+    sv = parse_symbolic_alt("1", 100, "T", "<%s>" % kind, info={"END": 105})
+    assert (sv.start, sv.end) == (100, 105)
+    assert (sv.affected_start, sv.affected_end, sv.length) == (101, 105, 5)
+    assert StructuralVariant.from_json(sv.to_json()) == sv
+
+
+@pytest.mark.parametrize("kind", ["DEL", "DUP", "INV", "CNV"])
+@pytest.mark.parametrize("info", [{}, {"END": 100}, {"END": 99}])
+def test_no_nonempty_span_cannot_delete_or_copy_the_anchor(kind, info):
+    with pytest.raises(ValueError, match="END greater than POS"):
+        parse_symbolic_alt("1", 100, "T", "<%s>" % kind, info=info)
+
+
+@pytest.mark.parametrize("reverse", [False, True])
+@pytest.mark.parametrize("kind,builder", [
+    ("DEL", _build_deletion_mutant_transcript),
+    ("DUP", _build_duplication_mutant_transcript),
+    ("INV", _build_inversion_mutant_transcript),
+])
+def test_exact_mutant_cdna_on_both_strands(kind, builder, reverse):
+    sequence = "ACGTTGCAAGCTTAGGCTAC"
+    tx = SimpleNamespace(exons=[SimpleNamespace(start=100, end=119)],
+                         start=100, end=119, sequence=sequence,
+                         on_backward_strand=reverse)
+    sv = parse_symbolic_alt("1", 109, "G", "<%s>" % kind, info={"END": 114})
+    a, b = (5, 10) if reverse else (10, 15)
+    if kind == "DEL":
+        expected = sequence[:a] + sequence[b:]
+    elif kind == "DUP":
+        expected = sequence[:b] + sequence[a:b] + sequence[b:]
+    else:
+        expected = sequence[:a] + reverse_complement(sequence[a:b]) + sequence[b:]
+    model = builder(sv, tx)
+    if kind == "INV":
+        # This builder preserves the layout but intentionally defers sequence.
+        observed = "".join(reverse_complement(sequence[s.start:s.end])
+                           if s.strand == "-" else sequence[s.start:s.end]
+                           for s in model.reference_segments)
+        assert observed == expected
+    else:
+        assert model.cdna_sequence == expected
+
+
+def test_vcf_deletion_does_not_remove_cftr_exon_terminal_anchor(tmp_path):
+    genome = cached_release(81)
+    tx = genome.transcript_by_id("ENST00000003084")
+    first, second = tx.exons[:2]
+    path = tmp_path / "anchor.vcf"
+    path.write_text(
+        "##fileformat=VCFv4.3\n"
+        "##INFO=<ID=END,Number=1,Type=Integer,Description=\"End\">\n"
+        "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\n"
+        "7\t%d\t.\tT\t<DEL>\t.\tPASS\tEND=%d\n" % (first.end, second.end))
+    (sv,) = load_vcf(str(path), genome=genome, parse_structural_variants=True)
+    effect = sv.effect_on_transcript(tx)
+    assert first not in effect.affected_exons
+    assert second in effect.affected_exons
+    a, b = first.end - first.start + 1, second.end - second.start + 1
+    assert effect.mutant_transcript.cdna_sequence == tx.sequence[:a] + tx.sequence[a+b:]
+
+
+def test_direct_constructor_span_defaults_remain_explicit():
+    sv = StructuralVariant("1", 100, "DEL", end=105)
+    assert (sv.affected_start, sv.affected_end, sv.length) == (100, 105, 6)

{varcode-9.2.0 → varcode-9.2.1}/varcode/structural_variant.py

@@ -179,8 +179,8 @@ class StructuralVariant(Variant):
 
     The SV position model:
 
-    * :attr:`start` — 1-based start of the affected region
-      (matches VCF POS).
+    * :attr:`start` — 1-based event/record position (matches VCF POS).
+      For parsed symbolic spans this is the retained padding base.
     * :attr:`end` — 1-based *inclusive* end. For a DEL/DUP/INV/CNV
       this is the SV endpoint on the same contig. For an INS it
       equals start (insertions are zero-width in reference coords).
@@ -188,8 +188,8 @@ class StructuralVariant(Variant):
       :attr:`mate_contig` / :attr:`mate_start`.
     * :attr:`affected_start` / :attr:`affected_end` — inclusive bases
       changed by a span event. These normally equal ``start`` / ``end``;
-      paired breakend events use them to exclude retained VCF padding
-      bases from exon and mutant-sequence annotation.
+      parsed symbolic spans and paired breakend events use them to exclude
+      retained VCF padding bases from exon and mutant-sequence annotation.
 
     Parameters
     ----------

{varcode-9.2.0 → varcode-9.2.1}/varcode/sv_allele_parser.py

@@ -198,10 +198,20 @@ def parse_symbolic_alt(
         if copy_number is not None:
             sv_info["copy_number"] = copy_number
 
+        # VCF symbolic POS is the retained padding base before the event.
+        # Keep record/junction coordinates unchanged; span consumers already
+        # use affected_start/end for paired breakends (#404).
+        affected_start = None
+        if top in {"DEL", "DUP", "INV", "CNV"}:
+            if int(end) <= int(start):
+                raise ValueError("A symbolic %s requires END greater than POS" % top)
+            affected_start = int(start) + 1
+
         return StructuralVariant(
             contig=contig,
             start=int(start),
             end=int(end),
+            affected_start=affected_start,
             sv_type=top,
             alt=alt,
             ref=ref or "N",

varcode-9.2.1/varcode/version.py

@@ -0,0 +1 @@
+__version__ = "9.2.1"

{varcode-9.2.0 → varcode-9.2.1/varcode.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: varcode
-Version: 9.2.0
+Version: 9.2.1
 Summary: Variant annotation in Python
 Author-email: Alex Rubinsteyn <alex.rubinsteyn@unc.edu>
 Project-URL: Homepage, https://github.com/openvax/varcode

{varcode-9.2.0 → varcode-9.2.1}/varcode.egg-info/SOURCES.txt

@@ -69,6 +69,7 @@ tests/test_string_helpers.py
 tests/test_structural_variant_annotator.py
 tests/test_structural_variants.py
 tests/test_symbolic_alleles.py
+tests/test_symbolic_sv_padding.py
 tests/test_timings.py
 tests/test_transcript_layout.py
 tests/test_transcript_model.py

varcode-9.2.0/varcode/version.py

@@ -1 +0,0 @@
-__version__ = "9.2.0"