API 参考

包入口

AnnData-native tools for spatial perturbation analysis.

class spatialperturb.PanelSpec(panel_id, figure_id, label, width_mm, height_mm, output_formats=('pdf', 'png', 'svg'), dpi=300, rasterize_dense=True, source_data_required=True, metadata=<factory>)

Publication panel output contract.

exception spatialperturb.SpatialPerturbSchemaError

Raised when an AnnData object does not satisfy the SpatialPerturb schema.

spatialperturb.aggregate_program_scores(adata, score_key, *, groupby)

Aggregate per-cell program scores into a tidy long table.

Return type:

DataFrame

spatialperturb.assign_perturbations(adata, *, barcode_columns=None, barcode_prefix=None, barcode_to_perturbation=None, min_counts=1.0, negative_label='unassigned', multiple_label='multiple', perturbation_col='perturbation', status_col='perturbation_status', include_col='include_for_inference', copy=False)

Assign a perturbation label to each cell from barcode feature counts.

Return type:

AnnData

spatialperturb.available_benchmarks()

Return available benchmark tracks and their required inputs.

Return type:

DataFrame

spatialperturb.available_datasets()

Return the bundled dataset catalog.

Return type:

DataFrame

spatialperturb.bootstrap_program_score_intervals(adata, score_key='program_scores', *, groupby=('cell_type', 'roi'), n_bootstrap=200, min_cells=50, seed=0)

Estimate bootstrap confidence intervals for grouped program means.

Return type:

DataFrame

spatialperturb.build_reference_programs(adata, *, control='control', groupby=None, perturbation_col='perturbation', status_col='perturbation_status', method='auto', sample_col=None, covariates=None, top_n=50, direction='both', min_cells_per_group=2, min_samples_per_group=2, effect_size_only=False, cell_type=None, roi=None, return_de_results=False)

Build per-perturbation reference programs from a Perturb-seq AnnData object.

Return type:

dict[str, list[str]] | tuple[dict[str, list[str]], DataFrame]

spatialperturb.build_signature_matrix(programs)

Convert a program-to-gene mapping into a binary membership matrix.

Return type:

DataFrame

spatialperturb.build_spatial_graph(adata, *, mode='knn', k=15, radius=None, spatial_key='spatial', graph_key=None, include_self=False, copy=False)

Construct a spatial neighbor graph and store it in adata.obsp.

Return type:

AnnData

spatialperturb.calibrate_program_scores(adata, programs=None, *, score_key='program_scores', groupby=('cell_type', 'roi'), n_random=100, n_label_shuffles=None, n_bins=20, min_cells=50, seed=0, layer=None, return_nulls=False)

Calibrate grouped program scores with expression-matched and label-shuffle nulls.

The returned table keeps every group, but marks groups with fewer than min_cells as exploratory so downstream reports do not overclaim tiny ROIs.

Return type:

DataFrame | tuple[DataFrame, DataFrame]

spatialperturb.collect_neighbors(adata, *, cells=None, graph_key=None, exclude_perturbed=True, status_col='perturbation_status', include_self=False, as_frame=False)

Collect neighbors for a set of cells from a stored spatial graph.

Return type:

dict[str, list[str]] | DataFrame

spatialperturb.compare_program_concordance(left, right)

Compare program definitions or program scores across two inputs.

Return type:

DataFrame

spatialperturb.dense_spatial_scatter(ax, coords, *, values=None, threshold=50000, force_rasterized=None, **kwargs)

Scatter spatial points with automatic rasterization for dense point clouds.

Return type:

PathCollection

spatialperturb.derive_perturbation_programs(de_results, *, group_col='perturbation', gene_col='gene', score_col='log2fc', top_n=50, direction='both')

Derive per-perturbation programs from a tidy DE table.

Return type:

dict[str, list[str]]

spatialperturb.differential_lr(adata, *, perturbation, control, graph_key=None, lr_network=None, groupby='perturbation', status_col='perturbation_status', source_groupby=None, target_groupby=None, layer=None, cell_type=None, roi=None)

Score ligand-receptor interactions between perturbed cells and their neighbors.

Return type:

DataFrame

spatialperturb.ensure_spatialperturb_schema(adata, *, metadata=None, copy=False)

Ensure required schema fields exist, filling defaults when needed.

Return type:

AnnData

spatialperturb.fetch_dataset(name, *, cache_dir, force=False)

Download and cache the raw public files for a registered dataset.

Return type:

dict[str, Any]

spatialperturb.from_tables(expression, *, obs=None, var=None, spatial=None, layers=None, metadata=None)

Build a schema-compliant AnnData object from tabular inputs.

Return type:

AnnData

spatialperturb.get_dataset_card(name)

Fetch a dataset card by name.

Return type:

DatasetCard

spatialperturb.intrinsic_de(adata, *, perturbation, control, groupby='perturbation', status_col='perturbation_status', features=None, layer=None, cell_type=None, roi=None, min_cells_per_group=2, min_samples_per_group=2, method='simple', sample_col=None, covariates=None, effect_size_only=False, min_cells=None)

Compare perturbed cells against control cells in the same ROI/cell type context.

Return type:

DataFrame

spatialperturb.load_demo_dataset(*, platform='spatial', paired=False, annotate=True, seed=0)

Return a deterministic demo dataset for tests, docs, and end-to-end examples.

Return type:

AnnData | tuple[AnnData, AnnData]

spatialperturb.load_public_dataset(name, *, cache_dir, prepared=True)

Load a prepared public dataset as an AnnData object.

Return type:

AnnData

spatialperturb.neighbor_de(adata, *, perturbation, control, graph_key=None, groupby='perturbation', status_col='perturbation_status', exclude_perturbed=True, drop_shared_neighbors=False, weight_by_distance=False, aggregate='mean', features=None, layer=None, cell_type=None, roi=None, min_cells_per_group=2, min_samples_per_group=2, method='simple', sample_col=None, covariates=None, effect_size_only=False, min_cells=None)

Compare neighbors around perturbed cells against neighbors around control cells.

Return type:

DataFrame

spatialperturb.neighbor_program_scores(adata, *, score_key='program_scores', graph_key=None, key_added='neighbor_program_scores', exclude_perturbed=True)

Average program scores across each cell’s neighborhood.

Return type:

DataFrame

spatialperturb.platform_concordance(left, right, *, group_col='perturbation', gene_col='gene', score_col='log2fc', top_n=50, level='both')

Compare per-perturbation gene effects between two result tables.

Return type:

DataFrame

spatialperturb.power_curve(adata, *, perturbation, control, mode='intrinsic', feature=None, sample_sizes=(5, 10, 20, 30), n_boot=100, alpha=0.05, groupby='perturbation', status_col='perturbation_status', graph_key=None, layer=None, method='simple', sample_col=None, cell_type=None, roi=None, random_state=0)

Estimate detection power across sample sizes by bootstrap resampling.

Return type:

DataFrame

spatialperturb.prepare_dataset(name, *, cache_dir, output_dir=None)

Standardize a public dataset into SpatialPerturb-compatible prepared outputs.

Return type:

dict[str, Any]

spatialperturb.program_redundancy_table(programs, *, threshold=0.5)

Return pairwise Jaccard overlap and coarse redundancy groups for programs.

Return type:

DataFrame

spatialperturb.qc_perturbations(adata, *, control, target_map=None, perturbation_col='perturbation', status_col='perturbation_status', layer=None, min_cells=5)

Summarize perturbation calling quality and target knockdown sanity checks.

Return type:

DataFrame

spatialperturb.read_stereoseq(path, **_)

Read a simple Stereo-seq-style cell-level export into AnnData.

Return type:

AnnData

spatialperturb.read_xenium(path, *, cell_group_path=None, roi_geojson_path=None, sample_name=None, load_molecules=False)

Read a simple Xenium-style cell-level export into AnnData.

Return type:

AnnData

spatialperturb.render_nature_methods_panels(report_dir, output_dir, *, strict=True)

Render fixed Nature Methods panel files from a completed report directory.

Return type:

dict[str, dict[str, str]]

spatialperturb.render_paper_figures(results, *, output_dir)

Render the canonical paper figures from benchmark result tables.

Return type:

dict[str, str]

spatialperturb.run_core_benchmark(dataset_or_adata, *, perturbations=None, control=None, target_map=None, lr_network=None, graph_key=None, sample_sizes=None, cell_type=None, roi=None, config=None, output_dir=None)

Run the core benchmark suite for a spatial perturbation dataset.

Return type:

dict[str, Any]

spatialperturb.run_cross_platform_benchmark(spatial_results, reference_results, *, config=None, output_dir=None)

Run the cross-platform concordance benchmark on paired DE result tables.

Return type:

DataFrame

spatialperturb.run_nature_methods_breast_analysis(spatial_input, *, reference_datasets=None, config=None, output_dir=None)

Run the publication-grade breast Xenium analysis for a Nature Methods Brief Communication.

Return type:

dict[str, Any]

spatialperturb.run_reference_projection_benchmark(spatial_input, *, reference_datasets, config=None, output_dir=None)

Project reference-derived programs onto a spatial or Xenium dataset.

Return type:

dict[str, Any]

spatialperturb.save_panel(fig, spec, source_data, output_dir, *, strict=True)

Save a production panel and its source data under a shared manifest.

Return type:

dict[str, str]

spatialperturb.score_programs(adata, programs, *, layer=None)

Score each program as the mean expression of its member genes per cell.

Return type:

DataFrame

spatialperturb.set_publication_rcparams()

Apply Nature-style editable-vector and small-panel Matplotlib defaults.

Return type:

None

spatialperturb.spatial_autocorrelation_scores(adata, *, score_key='program_scores', graph_key=None, n_permutations=100, seed=0)

Compute graph Moran-style spatial autocorrelation for program scores.

Return type:

DataFrame

spatialperturb.validate_reference_programs(reference_adata, *, programs=None, query_adata=None, control='control', perturbation_col='perturbation', status_col='perturbation_status', split_strategy='auto', top_n=50, direction='both', method='simple', seed=0, min_cells_per_group=2)

Validate reference programs by held-out recovery of perturbation labels.

Return type:

DataFrame

spatialperturb.validate_spatialperturb_schema(adata, *, require_graph=False)

Validate that an AnnData object follows the SpatialPerturb schema.

Return type:

None

I/O 与 Schema

AnnData I/O helpers for SpatialPerturb.

spatialperturb.io.from_tables(expression, *, obs=None, var=None, spatial=None, layers=None, metadata=None)

Build a schema-compliant AnnData object from tabular inputs.

Return type:

AnnData

spatialperturb.io.read_stereoseq(path, **_)

Read a simple Stereo-seq-style cell-level export into AnnData.

Return type:

AnnData

spatialperturb.io.read_xenium(path, *, cell_group_path=None, roi_geojson_path=None, sample_name=None, load_molecules=False)

Read a simple Xenium-style cell-level export into AnnData.

Return type:

AnnData

Schema helpers for AnnData objects consumed by SpatialPerturb.

exception spatialperturb.schema.SpatialPerturbSchemaError

Raised when an AnnData object does not satisfy the SpatialPerturb schema.

spatialperturb.schema.ensure_spatialperturb_schema(adata, *, metadata=None, copy=False)

Ensure required schema fields exist, filling defaults when needed.

Return type:

AnnData

spatialperturb.schema.schema_summary(adata)

Return a concise schema summary for validation reports.

Return type:

Series

spatialperturb.schema.validate_spatialperturb_schema(adata, *, require_graph=False)

Validate that an AnnData object follows the SpatialPerturb schema.

Return type:

None

预处理与图

Preprocessing and perturbation assignment utilities.

spatialperturb.pp.assign_perturbations(adata, *, barcode_columns=None, barcode_prefix=None, barcode_to_perturbation=None, min_counts=1.0, negative_label='unassigned', multiple_label='multiple', perturbation_col='perturbation', status_col='perturbation_status', include_col='include_for_inference', copy=False)

Assign a perturbation label to each cell from barcode feature counts.

Return type:

AnnData

spatialperturb.pp.qc_perturbations(adata, *, control, target_map=None, perturbation_col='perturbation', status_col='perturbation_status', layer=None, min_cells=5)

Summarize perturbation calling quality and target knockdown sanity checks.

Return type:

DataFrame

Spatial graph construction and neighborhood collection.

spatialperturb.gr.build_spatial_graph(adata, *, mode='knn', k=15, radius=None, spatial_key='spatial', graph_key=None, include_self=False, copy=False)

Construct a spatial neighbor graph and store it in adata.obsp.

Return type:

AnnData

spatialperturb.gr.collect_neighbors(adata, *, cells=None, graph_key=None, exclude_perturbed=True, status_col='perturbation_status', include_self=False, as_frame=False)

Collect neighbors for a set of cells from a stored spatial graph.

Return type:

dict[str, list[str]] | DataFrame

分析工具

Core analysis tools for spatial perturbation inference.

spatialperturb.tl.differential_lr(adata, *, perturbation, control, graph_key=None, lr_network=None, groupby='perturbation', status_col='perturbation_status', source_groupby=None, target_groupby=None, layer=None, cell_type=None, roi=None)

Score ligand-receptor interactions between perturbed cells and their neighbors.

Return type:

DataFrame

spatialperturb.tl.intrinsic_de(adata, *, perturbation, control, groupby='perturbation', status_col='perturbation_status', features=None, layer=None, cell_type=None, roi=None, min_cells_per_group=2, min_samples_per_group=2, method='simple', sample_col=None, covariates=None, effect_size_only=False, min_cells=None)

Compare perturbed cells against control cells in the same ROI/cell type context.

Return type:

DataFrame

spatialperturb.tl.neighbor_de(adata, *, perturbation, control, graph_key=None, groupby='perturbation', status_col='perturbation_status', exclude_perturbed=True, drop_shared_neighbors=False, weight_by_distance=False, aggregate='mean', features=None, layer=None, cell_type=None, roi=None, min_cells_per_group=2, min_samples_per_group=2, method='simple', sample_col=None, covariates=None, effect_size_only=False, min_cells=None)

Compare neighbors around perturbed cells against neighbors around control cells.

Return type:

DataFrame

spatialperturb.tl.platform_concordance(left, right, *, group_col='perturbation', gene_col='gene', score_col='log2fc', top_n=50, level='both')

Compare per-perturbation gene effects between two result tables.

Return type:

DataFrame

spatialperturb.tl.power_curve(adata, *, perturbation, control, mode='intrinsic', feature=None, sample_sizes=(5, 10, 20, 30), n_boot=100, alpha=0.05, groupby='perturbation', status_col='perturbation_status', graph_key=None, layer=None, method='simple', sample_col=None, cell_type=None, roi=None, random_state=0)

Estimate detection power across sample sizes by bootstrap resampling.

Return type:

DataFrame

Program and signature helpers for SpatialPerturb.

spatialperturb.signatures.aggregate_program_scores(adata, score_key, *, groupby)

Aggregate per-cell program scores into a tidy long table.

Return type:

DataFrame

spatialperturb.signatures.bootstrap_program_score_intervals(adata, score_key='program_scores', *, groupby=('cell_type', 'roi'), n_bootstrap=200, min_cells=50, seed=0)

Estimate bootstrap confidence intervals for grouped program means.

Return type:

DataFrame

spatialperturb.signatures.build_reference_programs(adata, *, control='control', groupby=None, perturbation_col='perturbation', status_col='perturbation_status', method='auto', sample_col=None, covariates=None, top_n=50, direction='both', min_cells_per_group=2, min_samples_per_group=2, effect_size_only=False, cell_type=None, roi=None, return_de_results=False)

Build per-perturbation reference programs from a Perturb-seq AnnData object.

Return type:

dict[str, list[str]] | tuple[dict[str, list[str]], DataFrame]

spatialperturb.signatures.build_signature_matrix(programs)

Convert a program-to-gene mapping into a binary membership matrix.

Return type:

DataFrame

spatialperturb.signatures.calibrate_program_scores(adata, programs=None, *, score_key='program_scores', groupby=('cell_type', 'roi'), n_random=100, n_label_shuffles=None, n_bins=20, min_cells=50, seed=0, layer=None, return_nulls=False)

Calibrate grouped program scores with expression-matched and label-shuffle nulls.

The returned table keeps every group, but marks groups with fewer than min_cells as exploratory so downstream reports do not overclaim tiny ROIs.

Return type:

DataFrame | tuple[DataFrame, DataFrame]

spatialperturb.signatures.compare_program_concordance(left, right)

Compare program definitions or program scores across two inputs.

Return type:

DataFrame

spatialperturb.signatures.derive_perturbation_programs(de_results, *, group_col='perturbation', gene_col='gene', score_col='log2fc', top_n=50, direction='both')

Derive per-perturbation programs from a tidy DE table.

Return type:

dict[str, list[str]]

spatialperturb.signatures.neighbor_program_scores(adata, *, score_key='program_scores', graph_key=None, key_added='neighbor_program_scores', exclude_perturbed=True)

Average program scores across each cell’s neighborhood.

Return type:

DataFrame

spatialperturb.signatures.program_redundancy_table(programs, *, threshold=0.5)

Return pairwise Jaccard overlap and coarse redundancy groups for programs.

Return type:

DataFrame

spatialperturb.signatures.score_programs(adata, programs, *, layer=None)

Score each program as the mean expression of its member genes per cell.

Return type:

DataFrame

spatialperturb.signatures.spatial_autocorrelation_scores(adata, *, score_key='program_scores', graph_key=None, n_permutations=100, seed=0)

Compute graph Moran-style spatial autocorrelation for program scores.

Return type:

DataFrame

spatialperturb.signatures.validate_reference_programs(reference_adata, *, programs=None, query_adata=None, control='control', perturbation_col='perturbation', status_col='perturbation_status', split_strategy='auto', top_n=50, direction='both', method='simple', seed=0, min_cells_per_group=2)

Validate reference programs by held-out recovery of perturbation labels.

Return type:

DataFrame

数据集与 benchmark

Dataset registry, lifecycle helpers, and demo data for SpatialPerturb.

class spatialperturb.datasets.DatasetCard(name, title, platform, accession, benchmark_track, source_url, description, status='metadata_only', raw_format='none', parser='demo', prepared_kind='adata', raw_archive_name=None, prepared_filename=None, matrix_filename=None, supports_auto_prepare=True, validation_note=None, supplementary_files=())

Metadata for a benchmarkable dataset.

spatialperturb.datasets.available_datasets()

Return the bundled dataset catalog.

Return type:

DataFrame

spatialperturb.datasets.fetch_dataset(name, *, cache_dir, force=False)

Download and cache the raw public files for a registered dataset.

Return type:

dict[str, Any]

spatialperturb.datasets.get_dataset_card(name)

Fetch a dataset card by name.

Return type:

DatasetCard

spatialperturb.datasets.load_demo_dataset(*, platform='spatial', paired=False, annotate=True, seed=0)

Return a deterministic demo dataset for tests, docs, and end-to-end examples.

Return type:

AnnData | tuple[AnnData, AnnData]

spatialperturb.datasets.load_public_dataset(name, *, cache_dir, prepared=True)

Load a prepared public dataset as an AnnData object.

Return type:

AnnData

spatialperturb.datasets.prepare_dataset(name, *, cache_dir, output_dir=None)

Standardize a public dataset into SpatialPerturb-compatible prepared outputs.

Return type:

dict[str, Any]

Benchmark orchestration for SpatialPerturb analyses.

spatialperturb.benchmarks.available_benchmarks()

Return available benchmark tracks and their required inputs.

Return type:

DataFrame

spatialperturb.benchmarks.run_core_benchmark(dataset_or_adata, *, perturbations=None, control=None, target_map=None, lr_network=None, graph_key=None, sample_sizes=None, cell_type=None, roi=None, config=None, output_dir=None)

Run the core benchmark suite for a spatial perturbation dataset.

Return type:

dict[str, Any]

spatialperturb.benchmarks.run_cross_platform_benchmark(spatial_results, reference_results, *, config=None, output_dir=None)

Run the cross-platform concordance benchmark on paired DE result tables.

Return type:

DataFrame

spatialperturb.benchmarks.run_nature_methods_breast_analysis(spatial_input, *, reference_datasets=None, config=None, output_dir=None)

Run the publication-grade breast Xenium analysis for a Nature Methods Brief Communication.

Return type:

dict[str, Any]

spatialperturb.benchmarks.run_reference_projection_benchmark(spatial_input, *, reference_datasets, config=None, output_dir=None)

Project reference-derived programs onto a spatial or Xenium dataset.

Return type:

dict[str, Any]

Paper-grade reporting and figure rendering helpers.

spatialperturb.reports.render_paper_figures(results, *, output_dir)

Render the canonical paper figures from benchmark result tables.

Return type:

dict[str, str]

投稿级 FigureKit

Production-grade panel rendering for publication workflows.

exception spatialperturb.figurekit.FigureKitError

Raised when a production figure contract is violated.

class spatialperturb.figurekit.PanelSpec(panel_id, figure_id, label, width_mm, height_mm, output_formats=('pdf', 'png', 'svg'), dpi=300, rasterize_dense=True, source_data_required=True, metadata=<factory>)

Publication panel output contract.

spatialperturb.figurekit.dense_spatial_scatter(ax, coords, *, values=None, threshold=50000, force_rasterized=None, **kwargs)

Scatter spatial points with automatic rasterization for dense point clouds.

Return type:

PathCollection

spatialperturb.figurekit.render_nature_methods_panels(report_dir, output_dir, *, strict=True)

Render fixed Nature Methods panel files from a completed report directory.

Return type:

dict[str, dict[str, str]]

spatialperturb.figurekit.save_panel(fig, spec, source_data, output_dir, *, strict=True)

Save a production panel and its source data under a shared manifest.

Return type:

dict[str, str]

spatialperturb.figurekit.set_publication_rcparams()

Apply Nature-style editable-vector and small-panel Matplotlib defaults.

Return type:

None

绘图与 CLI

Plotting helpers for SpatialPerturb result tables.

spatialperturb.pl.barcode_spread(adata, *, status_col='perturbation_status', ax=None)

Plot perturbation assignment status counts.

Return type:

Axes

spatialperturb.pl.lr_map(adata, lr_results, *, perturbation, graph_key=None, ligand=None, receptor=None, ax=None)

Plot perturbed cells and their neighbors colored by ligand and receptor expression.

Return type:

Axes

spatialperturb.pl.lr_pairs(lr_results, *, top_n=20, ax=None)

Plot top ligand-receptor differential scores.

Return type:

Axes

spatialperturb.pl.own_vs_neighbor(intrinsic_results, neighbor_results, *, perturbation=None, ax=None)

Plot intrinsic versus neighbor log fold changes for shared genes.

Return type:

Axes

spatialperturb.pl.platform_concordance(concordance_results, *, metric='spearman', ax=None)

Plot per-perturbation cross-platform concordance.

Return type:

Axes

spatialperturb.pl.power_curve(power_results, *, ax=None)

Plot empirical power as a function of sample size.

Return type:

Axes

Command-line interface for SpatialPerturb.

spatialperturb.cli.fetch_dataset_command(name, cache_dir=PosixPath('.spatialperturb-cache'), force=False)

Download raw public files for a registered dataset.

Return type:

None

spatialperturb.cli.list_benchmarks()

List benchmark tracks.

Return type:

None

spatialperturb.cli.list_datasets()

List built-in dataset cards.

Return type:

None

spatialperturb.cli.main()

Command group for SpatialPerturb utilities.

Return type:

None

spatialperturb.cli.prepare_dataset_command(name, cache_dir=PosixPath('.spatialperturb-cache'), output_dir=None)

Prepare a dataset into SpatialPerturb-compatible outputs.

Return type:

None

spatialperturb.cli.prepare_xenium_command(path, output_path, cell_group_path=None, roi_geojson_path=None, sample_name=None, load_molecules=False)

Prepare a Xenium directory into a schema-compliant h5ad file.

Return type:

None

spatialperturb.cli.render_paper_figures_command(dataset, cache_dir=PosixPath('.spatialperturb-cache'), output_dir=None, reference_dataset=None)

Render the canonical paper figures for a dataset.

Return type:

None

spatialperturb.cli.run_benchmark_command(dataset, cache_dir=PosixPath('.spatialperturb-cache'), output_dir=None, reference_dataset=None, method='auto')

Run the paper-grade benchmark workflow for a dataset.

Return type:

None

spatialperturb.cli.run_nature_methods_breast_analysis_command(spatial_input, output_dir, cache_dir=PosixPath('.spatialperturb-cache'), cell_group_path=None, roi_geojson_path=None, sample_name=None, reference_datasets='gse241115_breast_cropseq,gse281048_pathway_atlas', n_random=25, n_spatial_permutations=25, n_bootstrap=100, min_claim_cells=50)

Run the publication-grade Nature Methods breast short-communication workflow.

Return type:

None

spatialperturb.cli.run_reference_benchmark_command(spatial_input, output_dir, cache_dir=PosixPath('.spatialperturb-cache'), cell_group_path=None, roi_geojson_path=None, sample_name=None, reference_datasets='gse241115_breast_cropseq,gse281048_pathway_atlas')

Run the Xenium + Perturb-seq reference projection workflow.

Return type:

None

spatialperturb.cli.validate(path)

Validate that an h5ad file follows the SpatialPerturb schema.

Return type:

None

spatialperturb.cli.version()

Print the installed package version.

Return type:

None