Amptk

- improvement to `ufits filter` script to handle some of the filtering more gracefully and added a few more options
- added ability to supply a list of samples/barcodes to `ufits remove` and `ufits select` to make more flexible
- add new clustering module based on reference based clustering called `ufits cluster_ref`, it functions by quality trimming, dereplicating, chimera filtering, mapping to reference database. It can then also rescue unmapped reads and run de novo clustering on them followed by UTAX reference based clustering. I think that standard de novo clustering is superior to this approach but in some fringe cases it may be useful.

- slight modification to `ufits show` that also shows read lengths
- minor bug fix in `ufits taxonomy` that now properly exits if databases not installed
- changed the way `ufits filter` calculates index bleed from a sum of all counts per OTU, to now using the maximum value per sample.
- upgrade to `ufits meta` to also allow for splitting up data by taxonomy classification (if taxonomy from UFITS default method)

- added `ufits show` to count barcodes from demuxed data
- fixed bug in `ufits filter` if try to use `-s auto` option without providing a `--mock_barcode`

- bug fix where clustering log file was getting overwritten

- bug fix to `ufits filter` where the `--col_order` option now checks for samples before sorting to avoid error if sample passed but not in dataset, now it is ignored
- bug fix to `ufits filter` if using `-s` and `--keep_mock` options - script would die

- minor update to annotate mock community mapping to OTUs in `ufits filter`