a smarter way of handling repos of animal GEMs

[haowang-bioinfo]

This discussion was triggered by question of @johan-gson posted here:

• A number of GEM-type repos of animal species have been created and updated regularly using Human-GEM as template
• Would be good to reduce the manual work in curation and reviews, etc
• Suggestion ideas are welcome

[johan-gson]

Yeah, I just saw that there is a lot of manual work, reviews etc. for each species. It would probably be valuable to reduce that at some point.

[mihai-sysbio]

This was originally discussed when these repositories were instantiated. There are most likely pros and cons in either approaches - mono-repo vs separate repositories.

[johan-gson]

Perhaps reviews are only necessary for the first - the others are likely very similar?

[johan-gson]

But maybe that requires som automatic test cases or similar, I don't know if any such exist.

[mihai-sysbio]

It's envisioned that some automatic validation would be done via standard-GEM, but there is no plan yet for "test cases".

[haowang-bioinfo]

The design of template-based approach IS to reduce manual work in maintenance and curation from the very beginning.

To further reduce manual intervention and increase automation, would suggest:

• Continue to centralize routine functions into Human-GEM, on which animal GEM repos become thin clients and more dependant
• Develop GitHub actions to replace review works whenever possible

[haowang-bioinfo]

A generic and centralized animalGEM releasing function would probably be nice to include in addressing this and other related issues (#7 and #970). It could also evolve into GH actions run by bot later.

[mihai-sysbio]

I would propose an even more generic function called useAsTemplate(inputGEM).