scatrans.run_default_pipeline

scatrans.run_default_pipeline#

scatrans.run_default_pipeline(adata_input, groupby='condition', target_group='Disease', reference_group='Control', sample_col=None, organism='mouse', *, run_go_enrichment=True, gene_sets='GO_Biological_Process', filter_preset=None, show_plot=False)[source]#

End-to-end recommended workflow for first-time users.

Steps: active scoring → filter_active_genes → optional GO enrichment. Uses “Disease”/”Control” convenience defaults for target/reference.

The default for filter_preset is now auto-detected from the experimental design (via _resolve_simple_backend_kwargs): “pseudobulk” when sample_col is provided with >=3 samples per group (which triggers pseudobulk inside active_score_simple), otherwise “heuristic”. This keeps the thresholds consistent with the actual scale of active_score / unspliced_excess_residual (see WORKFLOW_PRESETS[“pseudobulk_report”]).

Returns a dict with keys:

  • adata, significant, all_results, candidates

  • enrichment (DataFrame or None)

  • filter_preset, backend (kwargs used for DE)

Parameters:

  • adata_input (Any)

  • groupby (str)

  • target_group (str)

  • reference_group (str)

  • sample_col (str | None)

  • organism (str)

  • run_go_enrichment (bool)

  • gene_sets (str)

  • filter_preset (str | None)

  • show_plot (bool)

Return type:

dict[str, Any]