rselect - select a random set of sequences from a FASTA file.
reps - extract exact k-word repeats based that occur in
sequences grouped in different clusters.
You'll need GHC or possibly another Haskell system, and the
Haskell bioinformatics library. The Makefile should work to
build and install (by default to your home directory) the
rselect [-r] n [m] input.seq
Selects n sequences from the file input.seq. If the optional
m is given, this limits the selection to happen only from the first
m sequences in the file, which may be more efficient. If -r is given,
the sequences will be reoriented randomly.
The selected sequences are written to standard output, so you
probably want to redirect them to a file.
reps k clustered.seq
Generate a list of repeated k-words (or k-grams) found in the sequences.
The sequences are expected to be on UniGene format, i.e. a FASTA file
with #-initiated comments separating the clusters.
A k-word is considered repeated if it is found in more than one of the
reps k clustering.lst sequences.seq
As above, but take a separate input clustering (and ignore any #'s in the
sequences. The clustering should consist of one line per cluster, with each
line containing the sequence identifier (first word after the initial '>'
in the FASTA header).
Do let me know about them, at <firstname.lastname@example.org>!