Add package

.github/workflows/build.yml

@@ -0,0 +1,35 @@
+name: build
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    env:
+      GITHUB_EVENT_NAME: ${{ github.event_name }}
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.12"]
+    timeout-minutes: 8
+
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+
+      - uses: actions/cache@v3
+        with:
+          path: ~/.cache/pre-commit
+          key: pre-commit-${{ runner.os }}-${{ hashFiles('.pre-commit-config.yaml') }}
+      - run: pip install laminci
+      - run: nox -s lint
+      - run: nox -s build

.gitignore

@@ -163,3 +163,4 @@ node_modules
 # lamin
 test.ipynb
 */test-perturbation
+lamin/data

noxfile.py

@@ -0,0 +1,22 @@
+import nox
+from laminci.nox import SYSTEM, build_docs, run, run_pre_commit, run_pytest
+
+nox.options.default_venv_backend = "none"
+
+
+@nox.session
+def lint(session: nox.Session) -> None:
+    run_pre_commit(session)
+
+
+@nox.session
+def build(session: nox.Session) -> None:
+    run(session, f"uv pip install {SYSTEM} .[dev]")
+    run_pytest(session, coverage=False)
+
+
+# Currently not enabled
+@nox.session
+def docs(session: nox.Session) -> None:
+    run(session, "lamin init --storage ./docsbuild --schema bionty,wetlab")
+    build_docs(session, strict=True)

pertpy_datasets/__init__.py

@@ -0,0 +1,7 @@
+__author__ = "Lukas Heumos"
+__email__ = "[email protected]"
+__version__ = "0.1.0"
+
+from pertpy_datasets.perturbation_curator import PerturbationCurator
+
+__all__ = ["PerturbationCurator"]

pertpy_datasets/perturbation_curator.py

@@ -0,0 +1,104 @@
+from typing import TYPE_CHECKING, Literal
+
+import anndata as ad
+
+if TYPE_CHECKING:
+    from lnschema_core import Record
+
+
+class _PerturbationValidatorUnavailable:
+    """Curator for perturbation data."""
+
+    def __init__(self):
+        raise RuntimeError("PerturbationValidator can only be instantiated if connected to a lamindb instance.")
+
+
+# Nested try because django might not be installed
+try:
+    from django.core.exceptions import ImproperlyConfigured
+
+    try:
+        import bionty as bt
+        import wetlab as wl
+        from cellxgene_lamin import CellxGeneFields, Curate
+        from lamin_utils import logger
+        from lamindb_setup.core.types import UPathStr
+        from lnschema_core.types import FieldAttr
+
+        class PerturbationCurator(Curate):
+            """Curator flow for Perturbation data - see pertpy-datasets."""
+
+            def __init__(
+                self,
+                adata: ad.AnnData | UPathStr,
+                var_index: FieldAttr = bt.Gene.ensembl_gene_id,
+                organism: Literal["human", "mouse"] = "human",
+                *,
+                verbosity: str = "hint",
+                cxg_schema_version: Literal["5.0.0", "5.1.0"] = "5.1.0",
+                using_key: str = "laminlabs/pertpy-datasets",
+            ):
+                """Curator flow for Perturbation data.
+
+                Args:
+                    adata: Path to or AnnData object to curate against the CELLxGENE schema.
+                    var_index: The registry field for mapping the ``.var`` index.
+                    categoricals: A dictionary mapping ``.obs.columns`` to a registry field.
+                        The PerturbationCurator maps against the required CELLxGENE fields and perturbation fields by default.
+                    organism: The organism name. CELLxGENE restricts it to 'human' and 'mouse' and therefore so do we.
+                    defaults: Default values that are set if columns or column values are missing.
+                    extra_sources: A dictionary mapping ``.obs.columns`` to Source records.
+                    verbosity: The verbosity level.
+                    cxg_schema_version: The CELLxGENE schema version to curate against.
+                    using_key: A reference LaminDB instance.
+                """
+                PT_DEFAULT_VALUES = CellxGeneFields.OBS_FIELD_DEFAULTS | {
+                    "cell_line": "unknown",
+                    "genetic_treatments": "",
+                    "compound_treatments": "",
+                    "environmental_treatments": "",
+                    "combination_treatments": "",
+                }
+
+                PT_CATEGORICALS = CellxGeneFields.OBS_FIELDS | {
+                    "cell_line": bt.CellLine.name,
+                    "genetic_treatments": wl.GeneticTreatment.name,
+                    "compound_treatments": wl.CompoundTreatment.name,
+                    "environmental_treatments": wl.EnvironmentalTreatment.name,
+                    "combination_treatments": wl.CombinationTreatment.name,
+                }
+
+                PT_SOURCES: dict[str, Record] = {
+                    "depmap_id": bt.Source.filter(name="depmap").one(),
+                    "cell_line": bt.Source.filter(name="depmap").one(),
+                    # "compound_treatments": bt.Source.filter(entity="Drug", name="chebi").first()
+                }
+
+                self.organism = organism
+
+                # Set the Compound source to chebi; we don't want output if the source has already been set
+                with logger.mute():
+                    chebi_source = bt.Source.filter(entity="Drug", name="chebi").first()
+                    wl.Compound.add_source(chebi_source)
+
+                super().__init__(
+                    adata=adata,
+                    var_index=var_index,
+                    categoricals=PT_CATEGORICALS,
+                    using_key=using_key,
+                    defaults=PT_DEFAULT_VALUES,
+                    verbosity=verbosity,
+                    organism=self.organism,
+                    extra_sources=PT_SOURCES,
+                    schema_version=cxg_schema_version,
+                )
+
+            def validate(self) -> bool:
+                """Validates the AnnData object against cellxgene and pertpy's requirements."""
+                return super().validate()
+
+    except ImproperlyConfigured:
+        PerturbationCurator = _PerturbationValidatorUnavailable  # type: ignore
+
+except ImportError:
+    PerturbationCurator = _PerturbationValidatorUnavailable  # type: ignore

pyproject.toml

@@ -1,3 +1,48 @@
+[build-system]
+build-backend = "hatchling.build"
+requires = ["hatchling"]
+
+[project]
+name = "pertpy_datasets"
+requires-python = ">=3.9, <=3.12"
+authors = [{name = "Lukas Heumos", email = "[email protected]"}]
+readme = "README.md"
+dynamic = ["version", "description"]
+classifiers = [
+    "License :: OSI Approved :: Apache Software License",
+    "Programming Language :: Python :: 3.9",
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+]
+dependencies = [
+    "cellxgene-lamin",
+    "findrefs",
+    "wetlab",
+    "requests",
+]
+
+[tool.hatch.version]
+path = "pertpy_datasets/__init__.py"
+
+
+[project.urls]
+Home = "https://lamin.ai/laminlabs/pertpy-datasets"
+
+[project.optional-dependencies]
+dev = [
+    "pre-commit",
+    "nox",
+    "pytest>=6.0",
+    "pytest-cov",
+    "nbproject_test",
+]
+
+[tool.pytest.ini_options]
+testpaths = [
+    "tests",
+]
+
 [tool.ruff]
 line-length = 120
 
@@ -62,6 +107,7 @@ convention = "google"
 "docs/*" = ["I"]
 "tests/*" = ["D"]
 "*/__init__.py" = ["F401"]
+"noxfile.py" = ["D"]
 
 [tool.mypy]
 strict = false