Skip to content
This repository has been archived by the owner on Oct 31, 2024. It is now read-only.

Create a card that lists all diseases related to the one being reviewed, and add to the overview display for a disease #241

Open
3 tasks
jmcmurry opened this issue Nov 4, 2019 · 7 comments
Assignees

Comments

@jmcmurry
Copy link
Member

jmcmurry commented Nov 4, 2019

From one disease see a preview of other diseases, whether they share:

  • phenotypes
  • causal gene
  • pathway

@monicacecilia monicacecilia added this to the Monarch-UI v1.1 milestone Nov 18, 2019
@monicacecilia monicacecilia changed the title Related diseases card (for disease overview) Create a card that lists all diseases related to the one being reviewed, and add to the overview display for a disease May 25, 2020
@jmcmurry
Copy link
Member Author

jmcmurry commented Jun 2, 2021

We discussed with @kshefchek today that this would be feasible to tackle; however it would be best if there were a cleaner API call to make.

  1. Kent to create some manual JSON (example being Marfan and Loyes-Dietz and similar)
  2. Kent to adjust the API
  3. In parallel with 2, Vince will create the component. This does not need to be embeddable.

@kshefchek
Copy link
Contributor

there was an outstanding question on whether to include implicit phenotypes in these counts @cmungall @pnrobinson your thoughts on this?

@pnrobinson
Copy link
Member

Implicit phenotypes should (nearly) always be counted by algorithms. Whether or not to show all of them in a UI is another question that "depends". Could you describe the use case?

@kshefchek
Copy link
Contributor

it would just be for showing these counts in the mock up above, perhaps also some hover over that shows the phenotypes (@jmcmurry ?) I like the idea of counting implicit phenotypes as we already support this via jaccard sim. I was more worried people might see a large phenotype intersection count and click on the page and see a much smaller number and be confused.

@pnrobinson
Copy link
Member

I mean that you may not want to show new nodes with implicit counts. I am not sure I understand the logic of the mockup, and it is not trivial to fine the best list of shared phenotypes for a larger list of diseases, that sounds like one of those nasty graph problems :-0.

@kshefchek
Copy link
Contributor

is this not just ranking jaccard sim scores? I may be under-thinking the problem

@pnrobinson
Copy link
Member

or me. We have probably reached the limits of what can easily be discussed on issues, let's discuss at a huddle or zoom!

@kshefchek kshefchek removed their assignment Dec 17, 2021
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

5 participants