Doc/Learn hub page or pages with summarized scientific "domain advice"

[jennaj]

Content to cover which tool to use, when, and why. A conversational tone seems best. Sort of a "Best of" for sci advice. Not a true tutorial, but could serve as a summary/baseline for one at a future time (or may already be included in a tutorial, but hard to find for quick Q&A).

• If opinion, state that.
• If differing opinions exist, state those as alternatives.
• If a tutorial includes that method (or compared methods), link in which one.
• If prior Q/A or method guides exists on a mailing lists or at Galaxy Biostars (https://biostar.usegalaxy.org) or Biostars (https://www.biostars.org) exist or other trusted public forum (google forums with tool author feedback and the like), link that in.
• If pubs, whitepapers, vetted opinion blogs, etc. exist that are appropriate to link in to help support the advice, include those. But I suggest not getting too hung up on this - domain expert advice can be subjective in many cases yet still valuable. Sci users can test out on own data.
  • side-by-side comparisons
  • considerations that touch on data content
  • usability with downstream tools (plus tips for usage to enable that, including important intermediate tools for prep and related)
  • performance/resource (e.g. tool X runs faster with less memory utilization)

Organize

• Put in http://galaxyproject.org Hub. /learn but link on /support? Not really a tutorial, but rather opinions to help with "what would a domain expert do" questions (exact quote from @pvanheus below).

Examples (RNA-seq):

• Hisat2 or Tophat?
• Featurecounts or Ht-seq count?

Suggested on Gitter by:

pvanheus @pvanheus Mar 31 07:28
these kind of workflow-y decisions is what GalaxyScientists was meant to discuss (and document) 
'cause I'm often asking those "what would a domain expert do" questions

[pvanheus]

And "is there an off the shelf, documented, community maintained workflow for X?"

[pvanheus]

just you and me, Jen?

[jennaj]

Maybe we start it then pull in others with expertise depending on the "domains" we cover / who is attending the hacks (or wants to help anyway, even remote)?

Thinking about the domains to target seems like a good next step. Would be nice to get to those without much organized existing help. Maybe can start list in this issue or create a new one and link it in?

Would also be good to draft hub and domain pages specific pages as markdown templates, ideally before the hack (can always tune it there).

And for this

pvanheus @pvanheus Apr 10 13:43
thanks @jennaj - trying to think how this could incorporate other tools...
esp. toolshed and biostars

Also, perhaps we can leverage the new https://galaxyproject.org/ search functionality in some way to make sure all resources are brought in and stay current without direct curation. But I haven't thought this through yet... :)

[pvanheus]

btw: https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-016-1457-z

I think this paper illustrates the overwhelming complexity that faces a RNA-seq practicioner well.

[jennaj]

Wow "Here, we evaluate 219 combinatorial implementations of the most commonly used analysis tools"

We probably won't cover them all :)