For a BAM file, find and mark all possible PCR duplicates. This is meant to be
used primarily with paired-end reads, since these have better resolution to
verify that we care seeing a legitimate PCR duplicate and not just reads that
happen to start at the same location.

The orientation for paired-end reads is assumed to be "FR" (forward-reverse).

Note: The BAM file must be sorted in order to find duplicates. For paired-end
      reads, the the proper-pair (0x4) flag must be set and the isize/tlen
      field must be correctly calculated.
Usage: bamutils pcrdup {options} infile.bam

Options:
    -frag                The reads are single-end fragments, so mark PCR
                         duplicated based only on the location of the read 
                         (not-recommended)

    -bam filename        Output BAM file with PCR duplicates marked

    -counts filename     Output of the number of reads at each position 

                         Note: this is actually the number of duplicate reads
                         at each position. If a position has multiple reads
                         mapped to it, but they are not pcr duplicates, then
                         there each will be reported separately.

    You must set either -bam or -counts (or both).