Change SSU search strategy

[abremges]

Currently, we use Barrnap (which is great and simple to use), but always assume domain Bacteria. This results in suboptimal annotations for e.g. Archaea. Either:

• Check other annotation results for most likely domain and adjust the command line call, or
• Use CheckM's ssuFinder module, which searches for all three domains and merges annotations.

I'm slightly leaning towards CheckM at the moment, which will also eliminate the dependencies on Barrnap and Bedtools, but this needs proper testing obviously!

[abremges]

Or go big and simply run Prokka on the full genome (w/ known domain from e.g. the conservative CheckM lineage)

[abremges]

Switched to checkm ssu_finder