search

LegumeFederation / DSCensor

Visualization and Organization for Assembly and Annotation Data
MIT License
0 stars 0 forks source link

Evaluate Tool Direction and Useability #26

Open ctcncgr opened 6 years ago

ctcncgr commented 6 years ago
  1. Do the views and statistics make sense?
  2. Are there obvious things missing that researchers will use?
  3. How should the tool let you access the data?
    • Should you be able to directly download?
    • How should bulk retrieval be handled if requested from multiple domains?
    • What would a search interface and it's response look like (Table, graph, icicle, combination, etc...)?
  4. How should the current views be accessed?
    • Main page with search and UI nav on top for specific type views? (Sensible?)
    • Land on genome type view, navigate from there?
    • Force type selection, then navigate from there?
  5. How would one get to this page?
    • How does it link from the new LegFed?
  6. How would one use the tool to do something interesting?
    • IGV
    • File Discovery Using Label(s), FileName(s) and Relationship(s)
    • Assembly and Annotation Global Stats Comparison
    • Linkouts
      • Cyverse (if available)
      • IGV
      • Other visualizers?
      • Intermine for the organism?
      • Intermine gene lists?
adf-ncgr commented 6 years ago

thanks Connor, this looks like a great set of questions to serve as a driver for the grand premiere. hope I did not seem too hesitant to put your work on display, I really just wanted to get people thinking about the question of how to put new stuff out in such a way that it has the maximum likelihood of having an impact...

ctcncgr commented 6 years ago

Not at all. I think there are still many questions.

In fact, I think I realized a major issue on the call today when we were talking about what it can display vs what it can search. I think the main interface needs to be able to display a variety of file types in some simple but interesting way to the user (easier said then done, but I think I have some ideas) that is searchable and takes advantage of the graph and the dynamic search aspects. The file type specific views are then selectable from the nav UI, and/or searchable based on type and/or taxa/whatever labels (but not a mix of file types). Currently I really only use the graph on the frontend to do the discovery for the IGV visualization and get the data to load the page, but the API can do much more than this.

I think this is a sensible direction.