monkeybread.stat.ligand_receptor_score

monkeybread.stat.ligand_receptor_score(adata, cell_to_neighbors, actual_scores, n_perms=100)

Calculates statistical significance of the co-expression of ligand-receptor pairs between neighboring cells.

Statistical test is as described in [HBB+22] (See Figure 4). This function runs this test separately on each ligand-receptor pair among a set of pairs provided by the user.

Parameters:

• adata (AnnData) – Annotated data matrix.

• cell_to_neighbors (Dict[str, Set[str]]) – A dictionary mapping each cell to its neighbors as calculated by monkeybread.calc.cell_neighbors()

• actual_scores (Dict[Tuple[str, str], float]) – The observed co-expression scores as calculated by monkeybread.calc.ligand_receptor_score()

• n_perms (Optional[int] (default: 100)) – Number of permutations to run in the permutation test

Return type:

Dict[Tuple[str, str], Tuple[ndarray, float]]

Returns:

: A mapping from ligand-receptor tuple pairs to a tuple containing the distribution of co-expression scores under permutation (i.e., null distribution) with the associated p-value of the observed co-expression score.