Amptk

- package has undergone a name change to reflect changes in the scripts. Originally the project started as essentially a wrapper for UPARSE and thus relied heavily on USEARCH. Coupled with originally supporting fungal ITS sequences, it was named UFITS (usearch fungal ITS). However, the current implementation of AMPtk relies very little on USEARCH and can support any amplicon based NGS dataset. Out of the box the following DB are packaged: fungal ITS, fungal LSU, 16S, insect/animal COI. Thus I feel that `amptk` is a better name that describes what the scripts do.
- option `-p, --pad` was added for `amptk ion`, `amptk illumina`, `amptk illumina2`, and `amptk 454` to allow user to turn off the padding with Ns to the `--trim_len`
- option `-c,--calculate` was added to `amptk filter` to control how the script calculates index-bleed. By default it calculates index-bleed into the mock community sample (`-b`) as well as out of the mock community into the rest of the samples. However, if members of the mock community are found in your samples, this calculated number is wrong, so if any members of your mock community are plausibly found in samples that you are sequencing, then you should use the `--calculate in` option.
- packaged databases had to be moved to a different sharing location (USDA now prevents use of dropbox), so they are now on Box, however it seems like the download speed is quite a bit slower. If anybody has recommendations for a free place to host these databases let me, need about 1 GB of space and need to be able to access with a directly link from the command line.

- move the mergereads function to general library
- better reporting for merge illumina reads for both `ufits illumina` `ufits illumina2`
- fix for `ufits illumina` to only require primer if amplicons are longer than the read length. This is to prevent amplicons that are shorter than the read length to be discarded as they are automatically trimmed/merged via `usearch -fastq_mergepairs` tool (and I can't change this). So the default behavior now is to require a forward primer via `--require_primer on` setting only if the amplicon length is longer than the read length. Read length is calculated automatically via sampling the first 50 reads, the automatic detection is overruled by the `--read_length` option

- fix critical bug in `ufits illumina` processing of reads where if reverse primer was not found read would be discarded

- update to `ufits taxonomy` to allow for taxonomy to be calculated elsewhere, pass the `-t, --taxonomy` option and a 2 column tsv file, OTU<tab>Taxonomy
- update to progress/multiprocessing steps
- re-write demultiplexing steps for faster processing
- support gzip files in `ufits illumina2`
- options in `ufits filter` for how the threshold is calculated for index-bleed filtering

- bug fix in `ufits illumina` where R2 reads were not getting trimmed correctly.

I bumped versions here to illustrate that UFITS has changed a little under the hood, now requires at least USEARCH v9.1.13 and requires at least VSEARCH v2.20. These changes were made to maximize speed and simplify the code. The scripts will terminate if they detect lower versions of both of these software tools. BIOM, RDP, Blast are still "soft dependencies".
- bug fix in `ufits taxonomy` where RDP taxonomy was not processed correctly for BIOM conversion
- fix for https://github.com/nextgenusfs/ufits/issues/14
- fix for https://github.com/nextgenusfs/ufits/issues/13 by moving to VSEARCH for this task
- fix for https://github.com/nextgenusfs/ufits/issues/12, now `ufits filter` requires you to add a mock community fasta file `--mc` if you specify a `-b, --barcode` to filter your data on
- fix for `ufits filter` to deal with OTU tables that have taxonomy already appended
- fix for `ufits cluster_ref` where script would die after conversion to VSEARCH as hard dependency
- re-write of `ufits heatmap` to have a few more options and more flexibility.
- update to docs as well as a section showing how to get your data into downstream statistical tools