P4 (WIP): GenCall-equivalent calling, BAF/LRR, sex, GTC v5, simurg genotype

- genotype/clusters.py: EGT aligned to manifest order (float32 mean R)
- genotype/gencall.py: flair, six-case likelihood, priors/posteriors,
  fuzzy sub-scores and score map, vectorised with the reference's
  float/double evaluation order and libm transcendentals
- genotype/baf_lrr.py, sex.py (rules v1/v2), sample.py (IDAT -> Gtc)
- pipeline/inputs.py shared by ingest and genotype
- pipeline/genotype.py + CLI: GTC per sample, sample_summary.tsv/json
  (OUT-060..063), run manifest
- GTCs from IDATs are byte-identical to the oracle for 36 GSA samples x
  2 normalization versions (NaN sign canonicalised to the x86 pattern)

Co-Authored-By: Claude Opus 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01814GKVzZVL3FuF3DeiZXhv

tests/oracle/test_p4_genotype.py

@@ -0,0 +1,59 @@
+"""Phase P4 acceptance on P0 data: GTCs produced from IDATs are byte-identical to oracle GTCs (gtc2vcf built without
+FMA) once the four free-text metadata fields (cluster/manifest file names, AutoCall date and version) are aligned.
+"""
+
+from __future__ import annotations
+
+import numpy as np
+import pytest
+from tests.conftest import GSA_BPM, GSA_EGT, GSA_IDATS, ORACLE_GTCS, ORACLE_GTCS_GT2, need
+
+from simurg.formats.bpm import read_bpm
+from simurg.formats.egt import read_egt
+from simurg.formats.gtc import encode_gtc, read_gtc
+from simurg.formats.idat import read_idat
+from simurg.genotype.clusters import align_clusters
+from simurg.genotype.sample import GenotypeParams, genotype_sample, manifest_context
+
+pytestmark = pytest.mark.data
+FREE_TEXT = ("cluster_file", "snp_manifest", "autocall_date", "autocall_version")
+NEG_NAN = np.frombuffer(np.uint32(0xFFC00000).tobytes(), dtype=np.float32)[0]
+
+
+def canonical_nans(g):  # type: ignore[no-untyped-def]
+    """NaNs computed by libm carry the platform's default sign (+ on aarch64, − on x86-64); SimurgArray always
+    writes 0xFFC00000 like the vendor's x86 build. Compare everything else byte for byte (known_unknowns K-014)."""
+    for k in ("b_allele_freqs", "logr_ratios", "genotype_scores"):
+        a = getattr(g, k).copy()
+        a[np.isnan(a)] = NEG_NAN
+        setattr(g, k, a)
+    return g
+
+
+@pytest.fixture(scope="module")
+def ctx():  # type: ignore[no-untyped-def]
+    need(GSA_BPM, GSA_EGT)
+    m = read_bpm(GSA_BPM)
+    return manifest_context(m, align_clusters(read_egt(GSA_EGT), m))
+
+
+@pytest.mark.parametrize(
+    ("folder", "norm"), [(ORACLE_GTCS, "1.2.0"), (ORACLE_GTCS_GT2, "1.1.2")], ids=["1.2.0", "1.1.2"]
+)
+def test_gtc_byte_identical(ctx, folder, norm) -> None:  # type: ignore[no-untyped-def]
+    need(folder, GSA_IDATS)
+    gtcs = sorted(folder.glob("*.gtc"))
+    assert len(gtcs) >= 18
+    for f in gtcs:
+        sid = f.stem
+        chip = sid.split("_")[0]
+        ref = read_gtc(f)
+        g = genotype_sample(
+            read_idat(GSA_IDATS / chip / f"{sid}_Grn.idat"),
+            read_idat(GSA_IDATS / chip / f"{sid}_Red.idat"),
+            ctx,
+            GenotypeParams(norm_version=norm),
+        )
+        for k in FREE_TEXT:
+            setattr(g, k, getattr(ref, k))
+        assert encode_gtc(g) == encode_gtc(canonical_nans(ref)), sid