Merge branch 'master' into GH1005-fix-cnv-fixed-denominator

adilraza99 · web-flow · commit 47583be51914 · 2026-03-03T17:27:40.000+05:30
diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
@@ -12,6 +12,10 @@ jobs:
       fail-fast: true
       matrix:
         python-version: ["3.10", "3.11", "3.12"]
+        numpy-spec:
+          # Keep this aligned with pyproject.toml: numpy = ">=2.0.2,<2.1"
+          - "==2.0.2"       # locked baseline
+          - ">=2.0.2,<2.1"  # latest allowed in declared range
     runs-on: ubuntu-latest
 
     steps:
@@ -23,8 +27,26 @@ jobs:
         with:
           python-version: ${{ matrix.python-version }}
 
-      - name: Verify NumPy version
-        run: poetry run python -c "import numpy; print('NumPy version:', numpy.__version__)"
+      - name: Install matrix NumPy version
+        run: poetry run pip install --upgrade --no-deps "numpy${{ matrix.numpy-spec }}"
+
+      - name: Verify NumPy version and spec
+        env:
+          NUMPY_SPEC: ${{ matrix.numpy-spec }}
+        run: |
+          poetry run python - <<'PY'
+          import os
+          import numpy
+          from packaging.specifiers import SpecifierSet
+
+          spec = SpecifierSet(os.environ["NUMPY_SPEC"])
+          version = numpy.__version__
+          if version not in spec:
+              raise RuntimeError(
+                  f"NumPy version {version} does not satisfy matrix spec {spec}"
+              )
+          print("NumPy version:", version, "| spec:", spec)
+          PY
 
       - name: Run unit tests
         run: poetry run pytest -v tests --ignore tests/integration --typeguard-packages=malariagen_data,malariagen_data.anoph
diff --git a/malariagen_data/anoph/cnv_data.py b/malariagen_data/anoph/cnv_data.py
@@ -296,7 +296,7 @@ def open_cnv_coverage_calls(
             marker = path + "/.zmetadata"
             if not self._fs.exists(marker):
                 raise ValueError(
-                    f"CNV coverage calls analysis f{analysis!r} not implemented for sample set {sample_set!r}"
+                    f"CNV coverage calls analysis {analysis!r} not implemented for sample set {sample_set!r}"
                 )
             store = _init_zarr_store(fs=self._fs, path=path)
             root = zarr.open_consolidated(store=store)
diff --git a/malariagen_data/anoph/h12.py b/malariagen_data/anoph/h12.py
@@ -558,6 +558,8 @@ def plot_h12_gwss_multi_overlay_track(
         show: gplt_params.show = True,
         x_range: Optional[gplt_params.x_range] = None,
         output_backend: gplt_params.output_backend = gplt_params.output_backend_default,
+        chunks: base_params.chunks = base_params.native_chunks,
+        inline_array: base_params.inline_array = base_params.inline_array_default,
     ) -> gplt_params.optional_figure:
         cohort_queries = self._setup_cohort_queries(
             cohorts=cohorts,
@@ -585,8 +587,11 @@ def plot_h12_gwss_multi_overlay_track(
                 min_cohort_size=min_cohort_size,
                 max_cohort_size=max_cohort_size,
                 sample_query=cohort_query,
+                sample_query_options=sample_query_options,
                 sample_sets=sample_sets,
                 random_seed=random_seed,
+                chunks=chunks,
+                inline_array=inline_array,
             )
 
         # Determine X axis range.
@@ -679,6 +684,8 @@ def plot_h12_gwss_multi_overlay(
         output_backend: gplt_params.output_backend = gplt_params.output_backend_default,
         gene_labels: Optional[gplt_params.gene_labels] = None,
         gene_labelset: Optional[gplt_params.gene_labelset] = None,
+        chunks: base_params.chunks = base_params.native_chunks,
+        inline_array: base_params.inline_array = base_params.inline_array_default,
     ) -> gplt_params.optional_figure:
         # Plot GWSS track.
         fig1 = self.plot_h12_gwss_multi_overlay_track(
@@ -700,6 +707,8 @@ def plot_h12_gwss_multi_overlay(
             height=track_height,
             show=False,
             output_backend=output_backend,
+            chunks=chunks,
+            inline_array=inline_array,
         )
 
         fig1.xaxis.visible = False
@@ -764,6 +773,8 @@ def plot_h12_gwss_multi_panel(
         output_backend: gplt_params.output_backend = gplt_params.output_backend_default,
         gene_labels: Optional[gplt_params.gene_labels] = None,
         gene_labelset: Optional[gplt_params.gene_labelset] = None,
+        chunks: base_params.chunks = base_params.native_chunks,
+        inline_array: base_params.inline_array = base_params.inline_array_default,
     ) -> gplt_params.optional_figure:
         cohort_queries = self._setup_cohort_queries(
             cohorts=cohorts,
@@ -789,6 +800,7 @@ def plot_h12_gwss_multi_panel(
                 window_size=window_size[cohort_label],
                 sample_sets=sample_sets,
                 sample_query=cohort_query,
+                sample_query_options=sample_query_options,
                 cohort_size=cohort_size,
                 min_cohort_size=min_cohort_size,
                 max_cohort_size=max_cohort_size,
@@ -799,6 +811,8 @@ def plot_h12_gwss_multi_panel(
                 height=track_height,
                 show=False,
                 output_backend=output_backend,
+                chunks=chunks,
+                inline_array=inline_array,
             )
             if i > 0:
                 track = self.plot_h12_gwss_track(x_range=figs[0].x_range, **params)
diff --git a/malariagen_data/veff.py b/malariagen_data/veff.py
@@ -1,6 +1,7 @@
 import collections
 import operator
 
+import pandas as pd
 from Bio.Seq import Seq  # type: ignore
 
 VariantEffect = collections.namedtuple(
@@ -62,7 +63,15 @@ def get_feature(self, feature_id):
         return self._idx_feature_id.loc[feature_id]
 
     def get_children(self, feature_id):
-        return self._idx_parent_id.loc[feature_id]
+        result = self._idx_parent_id.loc[feature_id]
+        # When there is only one child, pandas .loc returns a Series
+        # instead of a DataFrame. Ensure we always return a DataFrame
+        # so downstream code (e.g. .sort_values, column filtering) works.
+        if isinstance(result, pd.Series):
+            result = result.to_frame().T
+            # Preserve the index name from the parent DataFrame.
+            result.index.name = self._idx_parent_id.index.name
+        return result
 
     def get_ref_seq(self, chrom, start, stop):
         """Accepts 1-based coords."""
@@ -104,6 +113,17 @@ def get_effects(self, transcript, variants, progress=None):
         utr3 = list(children[children.type == "three_prime_UTR"].itertuples())
         introns = [(x.end + 1, y.start - 1) for x, y in zip(exons[:-1], exons[1:])]
 
+        # Guard: raise an informative error if the transcript has no CDS
+        # regions, as variant effect annotation is not meaningful for
+        # non-coding transcripts.
+        if len(cdss) == 0 and len(utr5) == 0 and len(utr3) == 0:
+            raise ValueError(
+                f"Transcript {transcript!r} has no CDS or UTR children. "
+                f"Variant effect annotation is only supported for "
+                f"protein-coding transcripts. This may indicate "
+                f"incomplete or incorrect genome annotations."
+            )
+
         effect_values = []
         impact_values = []
         ref_codon_values = []
diff --git a/tests/anoph/test_h12.py b/tests/anoph/test_h12.py
@@ -330,6 +330,36 @@ def test_h12_gwss_multi_with_analysis(fixture, api: AnophelesH12Analysis):
                 api.plot_h12_gwss_multi_panel(**params)
 
 
+@parametrize_with_cases("fixture,api", cases=".")
+def test_h12_gwss_multi_param_forwarding(fixture, api: AnophelesH12Analysis):
+    """Verify sample_query_options, chunks, and inline_array are
+    forwarded through multi-cohort H12 plotting functions."""
+    all_sample_sets = api.sample_sets()["sample_set"].to_list()
+    all_countries = api.sample_metadata()["country"].unique().tolist()
+    country1, country2 = random.sample(all_countries, 2)
+    cohort1_query = f"country == '{country1}'"
+    cohort2_query = f"country == '{country2}'"
+
+    h12_params = dict(
+        contig=random.choice(api.contigs),
+        sample_sets=all_sample_sets,
+        window_size=200,
+        min_cohort_size=1,
+        cohorts={"cohort1": cohort1_query, "cohort2": cohort2_query},
+        sample_query_options={"engine": "python"},
+        chunks="auto",
+        inline_array=False,
+    )
+
+    # Test multi-overlay — should not raise.
+    fig = api.plot_h12_gwss_multi_overlay(**h12_params, show=False)
+    assert isinstance(fig, bokeh.models.GridPlot)
+
+    # Test multi-panel — should not raise.
+    fig = api.plot_h12_gwss_multi_panel(**h12_params, show=False)
+    assert isinstance(fig, bokeh.models.GridPlot)
+
+
 def test_garud_h12_empty_window():
     import numpy as np
     from malariagen_data.anoph.h12 import _garud_h12

Original file line number	Diff line number	Diff line change
`@@ -296,7 +296,7 @@ def open_cnv_coverage_calls(`
`296`	`296`	`marker = path + "/.zmetadata"`
`297`	`297`	`if not self._fs.exists(marker):`
`298`	`298`	`raise ValueError(`
`299`		`- f"CNV coverage calls analysis f{analysis!r} not implemented for sample set {sample_set!r}"`
	`299`	`+ f"CNV coverage calls analysis {analysis!r} not implemented for sample set {sample_set!r}"`
`300`	`300`	`)`
`301`	`301`	`store = _init_zarr_store(fs=self._fs, path=path)`
`302`	`302`	`root = zarr.open_consolidated(store=store)`