Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Re-generate the .tsv files for the OBI synthetic data to include assessments / instruments and re-annotate them, and re-generate .jsonld files #196

Closed
1 task done
alyssadai opened this issue Nov 27, 2024 · 2 comments
Assignees

Comments

@alyssadai
Copy link
Contributor

alyssadai commented Nov 27, 2024

  • share the .jsonld files with the OBI and make sure that they can load and then find these datasets
@alyssadai alyssadai mentioned this issue Nov 27, 2024
9 tasks
@alyssadai alyssadai self-assigned this Nov 27, 2024
@alyssadai alyssadai moved this to Implement - Active in Neurobagel Nov 27, 2024
@alyssadai
Copy link
Contributor Author

alyssadai commented Nov 27, 2024

For reviewer:

In total, in terms of instruments, we were able to annotate using Cognitive Atlas (see full lists here https://docs.google.com/spreadsheets/d/1kMsnNYk3D8rNWRiBr94Aiv72q8rREjIl418qKRQxDvo/edit?usp=sharing):

  • 4/30 in CAN-BIND
  • 0/14 in POND (but this was because of their unusual instruments, which were like categories rather than actual concrete assessments)
  • 5/34 in ONDRI
    • One weird thing was they had UPDRS with 2 different names

Some other notes on the process, mostly re: annotation tool:

  • Needing to re-annotate the entire dataset from scratch was a bit annoying (issue opened)
  • Some aspects of the assessment tool annotation are still a little non-user-friendly, particularly (issue opened):
    • User first identifies all columns in their TSV about any assessment, and then matches them to specific assessments from CogAtlas. However, if assessments are not found in CogAtlas, user has to go back up to the initial column categorization and remember & de-select all of the ones that they couldn't annotate
    • Not being able to search by instrument abbreviations
    • The UI for the assessment-to-column matching is a little "bulky", in that when verifying columns for each assessment, I have to scroll through the entire list including the ones I have annotated as being about other assessments
      • Some initial ideas: maybe we could use a two-column layout for the column names, such that the ones annotated with the currently selected assessment are all grouped together? or otherwise visually group them somehow
  • Looks like the annotation tool is not fully refreshed when you go back to "Home" and upload a new file for annotation - the same assessment tools as in the previous round are selected by default (issue opened)
    • An "annotate another dataset" button would be nice!
  • The age transformation annotation is easy to miss - even though I encountered the nice warning message on the download page about age missing "annotations", it took me several seconds to see on the age annotation page that what I was missing was the transformation (issue opened)

@alyssadai alyssadai moved this from Implement - Active to Implement - Done in Neurobagel Nov 28, 2024
@surchs surchs moved this from Implement - Done to Review - Active in Neurobagel Nov 29, 2024
@surchs
Copy link
Contributor

surchs commented Dec 2, 2024

Thanks @alyssadai, these are all very good points about the annotation tool. I think it'd make sense to discuss them even before the bigger changes we have planned and see what could be quickly addressed.

🧑‍🍳

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
Status: Review - Done
Development

No branches or pull requests

2 participants