Question about using results of addon in funannotate2 annotate

[Alteroldis]

Dear Dr. Palmer, I'm getting to know funannotate2 and so far I have positive impressions of the speed.

1. I suggest adding variables to the installation manual like this:
   export AUGUSTUS_CONFIG_PATH=path-to-augustus/config
   PATH=$PATH:path-to-GeneMark-ETP/bin/gmes:path-to-augustus/bin:path-to-augustus/scripts/:path-to-GlimmerHMM/bin/:path-to-GlimmerHMM/train/:
2. Did I understand correctly that the result of the addons' work should be added via the --curated-names option of funannotate2 annotate? I suggest adding link to this project to 'Functional Annotation' section of documentation.
3. I don't quite understand what the difference is between annotations for interpro via funannotate2 annotate and this addons?

[nextgenusfs]

It's mostly about maintenance as well as these tools in f2a are too large or have too many dependencies to be wrapped into funannotate2.

Separating out these functional annotation jobs that have large database and/or ever-changing formats, leaves intact f2 functionality. Many of the bugs that users report in f1 are related to parsing these tools. So I don't really have time to help everyone with this all the time, so my expectation for this repo is that it is more community driven. f2 is setup to take any 3 column annotation file and note that several of these tools aren't quite working yet, hence I haven't tagged a release. But the idea is that you run f2 train, than f2 predict to generate the primary gene model annotations. You could then run f2a [optional] and then finally f2 annotate to pull together all the functional annotation. Making it somewhat modular also makes it easier for users to run jobs in parallel on their own computer systems.