Phamerate

Made the Roary-style pangenome analyses optional.

**Full Changelog**: https://github.com/chg60/phamerate/compare/v0.2.1...v0.2.4

**Full Changelog**: https://github.com/chg60/phamerate/compare/v0.2.0...v0.2.2

Fixed a rare (per our use case...) but important bug.

If inputs are Genbank flat files where CDS features do not have both a protein_id and a locus_tag, the FASTA header constructed for pham assembly would begin with a space and MMseqs2 would fail to run. Fixed by utilizing either a locus_tag or a protein_id in the header. Protein-coding genes with neither qualifier will be skipped over (and print message so this is not hidden).

In this release:
* Added support for Genbank flat files as inputs (inputs may be mixed - some FASTA & some Genbank). CDS features are parsed from flat files with the locus_tag, protein_id, and product qualifiers combined to form a (hopefully) unique header for each sequence. Should theoretically work for all prokaryotic genomes that come from RefSeq.

In this release:
* updated README.md

In this release:
* Improved documentation
* Added some Roary-style output files (gene_presence_absence.csv, summary_statistics.txt)
* Added simple tab-delimited output file mapping resultant phamids to the genomes they are found in