Butina

Implementation of the Butina clustering algorithm.

ButinaClustering(params)

Bases: Clusterer[ButinaParams]

Butina clustering algorithm.

Initialize the Butina clustering algorithm.

cluster(sims)

Cluster the pass's compounds with RDKit Butina.

PARAMETER	DESCRIPTION
sims	Output of compute_pairwise over the compounds to cluster. TYPE: TanimotoSimilarities

RETURNS	DESCRIPTION
ClusterResult	A ClusterResult keyed by sims.ids.

ButinaParams(similarity_cutoff=0.7, sparse_above=None)

Bases: ClusterParams

Parameters for the Butina clustering algorithm.

method = ClusterMethod.BUTINA

The clustering method used which is Butina.

similarity_cutoff = 0.7

The tanimoto similarity cutoff for clustering where higher values indicate more similar compounds. Must be between 0 and 1.

sparse_above = None

The largest cluster size above which the sparse (edge-based) path is used instead of the dense RDKit matrix. If None, the dense path is always used.

__post_init__()