kclust - Perform fast and sensitive clustering

kClust(sequences, clustering_thresh=30, create_alignments=False)

Uses kClust to generate clusters of amino acid sequences.

Parameters:
  • sequences (ost.seq.SequenceList) – All sequences you want to cluster.

  • clustering_thres – Sequence identity threshold to build clusters. Note, that clustering_thresh is more a rule of thumb, since compositional bias in the sequence can also play a role. The value gets transformed in a bitscore, that is used as an input parameter of kClust

  • create_alignments – Flag, wether the alignments of the clusters get calculated. Requires clustalw in the path.

Returns:

A list of cluster instances

Raises:

FileNotFound if kClust could not be located.

class cluster(sequences, representative_id)

Holds the information of one cluster

sequences

SequenceList containing all sequences of the cluster

representative_id

a string, that contains the name of the representative sequence of the cluster as estimated by kClust.

alignment

alignment handle containing a multiple sequence alignment of all sequences of the cluster. Gets only calculated when binding is called with appropriate flag.