Python interface to the cd-hit clustering program. Given a collection of redundant sequence records, (an iterable of (header, sequence) tuples) and a sequence identity threshold, the filter function returns an iterable of non-redundant records:
>>> import pcdhit >>> filtered_records = pcdhit.filter(records, thr=0.9)
In practice, the function dumps the redundant records in a fasta alignment, run cd-hit:
$ cdhit -i <redundant_fasta> -o <non_redundant_fasta> -c <threshold>
via subprocess.Popen and parse the non-redundant records from the cd-hit output file.