Finds regions of unusual deletions (CLIP-seq)

Given a set of BAM files, we search for areas where there are an unusual
amount of deletions.  For CLIP-Seq, this can be an indicator of the location
of protein-RNA interaction.

Output is either a BED file or a FASTA format file containing these hotspots.
Only unique regions are returned across all files.

See: Zhang and Darnell, Nature Biotechnology (2011)
     doi:10.1038/nbt.1873
     pmid:21633356


Usage: bamutils cims {opts} in.bam {in.bam...}

Options:
    -fasta ref.fa    Ouput in FASTA format (requires reference genome.fa)
                     [default: BED output]

    -flanking N      The number of flanking bases on either side to report
                     (FASTA output only) [default: 12]

    -cutoff N        Cut-off % for deletions - if the % of reads that
                     include a deletion at a position is higher than this
                     number, the fragment is reported (0->1.0)
                     [default: 0.1]

    -ns              Don't take the strand of the read into account

    -window N        The maximum length of a deletion window
                     [default: 20]