Consider adding annotated contigs as main input to the pipeline about funcscan HOT 3 CLOSED

We've hit a problem with optionally including annotated contigs as input.

I've actually got this working, however the code has got quite messy now. But this has also led to two issues which makes me feel like I would rather not support this:

1. DeepARG takes as input specifically nucleotide fasta out of PROKKA. Therefore if we don't run PROKKA on those samples that are supplied with the annotation fasta file, then those samples will not be executed by PROKKA. I.e. we would always have to run PROKKA on everything anyway regardless
2. After thinking about it, I am also now concerned about version and software consistency. By allowing users supply their own FAA files, this could be from versions different from what we have in the pipeline, and in fact could be from different tools that may not be supported by other steps of our pipeline.

Therefore I would actually rather not support this. If you're concerned about speed, you could just Prodigal instead.

What do you think about this?

from funcscan.

Where I got to: 8bf973f

from funcscan.

I got an OK to skip this from Martin, and with the addition of Prodigal this should massively speed up the process anyway of those who are concerned about run time.

from funcscan.