mantispy.tl.pathway_coherence

mantispy.tl.pathway_coherence(adata, net, gene_key='Metadata_Gene', metric='cosine', use_rep=None, min_genes=5, n_permutations=1000, seed=0, key_added='pathway_coherence', copy=False)

Score how similar the profiles of each gene set’s genes are.

For each set, the mean pairwise similarity among the profiles of its genes is compared with random sets of the same number of profiles. This checks whether a genetic screen recovers known biology.

Parameters:

• adata (AnnData) – One profile per gene, normally the output of consensus().

• net (DataFrame) – A gene-set network from gene_sets(), or any frame with source and target.

• gene_key (str (default: 'Metadata_Gene')) – obs column holding the gene symbol.

• metric (str (default: 'cosine')) – Similarity between profiles, "cosine" or "pearson".

• use_rep (str | None (default: None)) – Measure in obsm[use_rep] instead of X.

• min_genes (int (default: 5)) – Sets with fewer of their genes present in the screen are skipped.

• n_permutations (int (default: 1000)) – Number of random sets in the null.

• seed (int (default: 0)) – Seed for drawing the random sets.

• key_added (str (default: 'pathway_coherence')) – Name for the output table.

• copy (bool (default: False)) – Return a modified copy instead of mutating in place.

Return type:

AnnData | None

Returns:

None, or the modified copy. Writes uns["mantispy"][key_added] with set, n_genes, coherence, pvalue and qvalue, sorted by coherence.

Raises:

KeyError – obs has no column gene_key.

Notes

Rank sets by coherence rather than by p-value. With 1000 permutations every strongly coherent set reaches the p-value floor of 1/1001 and ties there. The p-value shows whether a set is coherent, and the coherence orders the sets that are.