grassp.tl.markov_clustering

markov_clustering(adata, resolution=1.2, key_added='mc_cluster')

Run Markov Clustering (MCL) on the neighbour graph.

The algorithm operates on the connectivities matrix written by scanpy.pp.neighbors() and assigns a cluster label to every observation. Labels are stored as a categorical column in adata.obs[key_added].

Parameters:

adata AnnData

Annotated data matrix which already contains adata.obsp['connectivities'].

resolution float (default: 1.2)

Inflation parameter of the MCL algorithm. Larger values yield more, smaller clusters. Typical range: 1.2 – 5.

key_added str (default: 'mc_cluster')

Observation key used to store the resulting cluster labels (default "mc_cluster").

Returns:

None The function modifies adata in place.