scatrans.extract_gene_lists

scatrans.extract_gene_lists#

scatrans.extract_gene_lists(de_results, *, logfc_cutoff=0.5, pval_cutoff=0.05, logfc_direction='up', separate_directions=False, name_prefix=None, gene_case=None)[source]#

Extract named gene lists from one or more DE result DataFrames for downstream enrichment.

This is the recommended way to prepare inputs for compare_enrichment when you have results from differential_expression() or the all_results from active_score().

Supports up / down / both, and can split directions into separate named sets (e.g. “GA_up”, “GA_down”) so you can enrich and visualize them distinctly (useful for upset plots and grouped dotplots).

Parameters:

  • de_results (DataFrame or dict[str, DataFrame]) –

    • Single DataFrame: treated as one unnamed contrast (you can use name_prefix).

    • dict {contrast_name: de_df}: each df is processed and keys become the cluster names.

  • logfc_cutoff (float) – Minimum |logFC| (sign handled by direction).

  • pval_cutoff (float) – Max p_adj (or p_val if no p_adj column).

  • logfc_direction ({"up", "down", "both"})

  • separate_directions (bool) – If True and direction in {“both”, “up”, “down”} context, will emit separate entries “<name>_up” and “<name>_down”. Great for “up vs down” enrichment comparison.

  • name_prefix (str, optional) – Prepended to generated names when a single df is passed.

  • gene_case (optional) – Passed through to gene cleaning.

Returns:

Ready to pass to compare_enrichment(gene_clusters=…).

Return type:

dict[str, list[str]]

Example

# Multiple contrasts de_dict = {“GA_vs_Ctrl”: ga_res, “GB_vs_Ctrl”: gb_res} gene_sets = scat.extract_gene_lists(

de_dict, logfc_cutoff=0.5, pval_cutoff=0.05, logfc_direction=”up”

) comp = scat.compare_enrichment(gene_sets, gene_sets=”GO_Biological_Process”, organism=”mouse”)

# Up and down as separate “clusters” for upset / grouped dotplot gene_sets = scat.extract_gene_lists(

de_dict, logfc_direction=”both”, separate_directions=True

)