scHopfield.tools.get_network_links

scHopfield.tools.get_network_links(adata: AnnData, cluster_key: str = 'cell_type', return_format: str = 'dict') → Dict[str, DataFrame] | DataFrame

Extract network links (edges) from interaction matrices.

Converts interaction matrices (W) into edge list format suitable for network analysis. Each edge represents a gene-gene interaction with its coefficient value.

Parameters:

• adata (AnnData) – Annotated data object with fitted interaction matrices

• cluster_key (str, optional (default: 'cell_type')) – Key in adata.obs for cluster labels

• return_format (str, optional (default: 'dict')) – Return format: ‘dict’ returns dict of DataFrames per cluster, ‘combined’ returns single DataFrame with cluster column

Returns:

If ‘dict’: Dictionary mapping cluster names to DataFrames with columns:

• source: source gene

• target: target gene

• coef_mean: interaction coefficient

• coef_abs: absolute value of coefficient

If ‘combined’: Single DataFrame with additional ‘cluster’ column

Return type:

dict or pd.DataFrame