mira.tl.summarize_mutual_information#
- mira.tl.summarize_mutual_information(adata1, adata2, *, key='X_topic_compositions')#
Calculate the total mutual information between expression and accessibility topics. A value of 0 indicates low correspondance between modes, while 0.5 indicates high correspondance. Good models for cell systems should have high mutual information.
- Parameters
- expr_adataanndata.AnnData
AnnData object with expression features, must have “X_topic_compositions” in .obsm.
- atac_adataanndata.AnnData
AnnData object with accessibility features, must have “X_topic_compositions” in .obsm.
- Returns
- mutual informationfloat
Total mutual information between expression and accessibility topics. Use this metric to evaluate the concordance between models across modes.