Python function to plot relative influence for each spatially adjacent cell type.

The influence of cell types positioned next to a given velocity cluster, where influence Cell types with higher ligand expression will contribute more to the final sum, resulting in a weighted influence score. This calculation is run based on tangram-sc transferred labels for velocity clusters and cell type clusters.

Usage

plot_influence_proportions(
  adata,
  adata_st,
  celltype_key,
  tangram_celltype_column,
  velocity_cluster_key = "velocity_clusters",
  tangram_result_column = "velocity_label_transfer",
  color = None,
  colormap = None,
  top_n_ligands = 5,
  save = "cell_influence_proportions.png",
  dpi = 300
)

Arguments

adata: An anndata object that has had get_velocity_ligands_spatial run on it.
adata_st: An anndata object containing spatially resolved data.
celltype_key: Column name in adata.obs denoting sender cell type labels to be visualized.
tangram_celltype_column: Column name in adata_st.obs denoting cell type labels that have been label-transferred from adata.obs. If not given, this function will run tangram-sc to transfer labels.
velocity_cluster_key: Default "velocity_clusters". Name of the column in adata.obs where velocity cluster labels are stored.
tangram_result_column: Default "velocity_label_transfer". Name of the column in adata_st.obs where velocity cluster labels are stored.
color: Argument for matplotlib plotting backend.
colormap: Argument for matplotlib plotting backend.
top_n_ligands: Default 5. The top n ligands (ranked by feature importance) to consider.
save: Default "cell_influence_proportions.png". Filename to save plot as.
dpi: Dpi for saving.

Value

A matplotlib axes object. Also saves a plot in working directory with filename given by save