Scanpy Tools

Source: src/single_cell_python_tools/tools/_functions_4_scanpy.py

These helpers are exported through sctl.tl.

General Utilities

  • print_size_in_MB(obj): print object size in megabytes.

  • df_loadings_ordered_byPC(adata, ascending=True, save_table=False, output_dir="./", output_prefix="adata"): return PCA loadings ordered by principal component.

  • cef_to_adata(data_dir, data_prefix, n_obs=None, n_skiprows=0, cef_delimiter_tab=True, save_to_h5ad=False): create an AnnData object from CEF-style inputs.

Annotation Helpers

  • annotate_marker_genes(adata, gene_names, min_n_counts, obs_key): annotate cells with marker-gene labels.

  • label_cells_by_single_gene_expression(adata, gene_name1, min_n_counts1=0, use_raw=True, use_percentile=False): label cells by one gene.

  • label_cells_by_double_gene_expression(adata, gene_name1, gene_name2, min_n_counts1=0, min_n_counts2=0, use_raw=True, use_percentile=False): label cells by two genes.

Differential Expression And Enrichment

  • rank_genes(adata, output_dir="./adata_output/", output_prefix="adata_", save_output=True, wilcox=True, logreg=True, t_test=True, rank_use_raw=False, obs_key="leiden", n_jobs=1, **parameters): run Scanpy rank-gene tests and optionally write tables.

  • rank_genes_obscat1_vs_obscat2(adata, output_dir="./adata_output/", output_prefix="adata_", save_output=True, wilcox=True, logreg=True, t_test=True, rank_use_raw=False, n_jobs=1, obs_key="leiden", obscat1=None, obscat2=None, **parameters): rank genes between two categories.

  • diff_exp(adata, groupby, group1, group2, layer=None): compare expression between two groups.

  • GSEA_enrichr_all_clusters(output_dir="./adata_output/", output_prefix="adata_", test_library_names=None, top_nth=0.10, n_jobs=1, **parameters): run Enrichr analysis from rank-gene output tables.

Filtering And Summaries

  • filter_obs(data, var, func=None): filter observations in place using .obs columns or expression values.

  • average_feature_expression(adata, groupby_key, layer=None, use_raw=False, log1p=False, zscore=False, subtract_mean=False): average feature expression by group.

  • average_obs_feature_per_group(adata, groupby_key, obs_keys): average observation-level features by group.

  • make_df_obs_adataX(adata, layer=None, index=True, varcolumns=None, include_obs=True, obscolumns=None, use_raw=False): build a pandas.DataFrame from adata.X, adata.layers, or adata.raw.

Notes

The current tools/__init__.py exports _functions_4_scanpy.py. The _ingest_verbose.py module exists separately and is documented on Ingest Verbose.