Says 0 function associations even if there are some about monarch-ui HOT 14 CLOSED

@kshefchek thanks for point that out.

Looks like a bug in Ontobio. Will issue a PR for it soon.

from monarch-ui.

I think the TODO here is to make a ticket in biolink and link it here.

from monarch-ui.

GOlr (golr-aux) has 10 function associations: https://api.monarchinitiative.org/api/bioentity/gene/HGNC%3A17158/function?rows=100&unselect_evidence=false&exclude_automatic_assertions=false&fetch_objects=true&use_compact_associations=false

Association counts for HGNC:17158 yields the following:

  "association_counts": {
    "interaction": 47,
    "homolog": 20,
    "anatomy": 20,
    "pathway": 12,
    "function": 9,
    "phenotype": 9,
    "disease": 1,
    "publication": 55,
    "variant": 3,
    "ortholog-interaction": 119,
    "ortholog-anatomy": 56,
    "ortholog-function": 22,
    "ortholog-phenotype": 26,
    "ortholog-pathway": 9
  },

There is a slight mismatch between Monarch Solr and GOlr.
But this still does not explain the value of '0 function associations' shown on that page.

The issue might be two fold:

1. Mismatch between Monarch Solr and GOlr
2. A bug in the UI code that doesn't show the proper counts

from monarch-ui.

Following the network calls on that page, the data comes from https://api-dev.monarchinitiative.org/api/bioentity/gene/HGNC:17158/function?fetch_objects=true&start=0&rows=25

where:
"numFound": 0,

from monarch-ui.

This looks to be fixed but theres still some oddities on this page that could go into a separate ticket or be tracked here:

• We're saying there are 9 distinct go terms but I only see 7
• Genes are linked to uniprot IDs that don't resolve to blank pages. Do we need protein pages - or have these forward to the appropriate gene page?

from monarch-ui.

@kshefchek Great to hear!

I agree with you, let's close this issue.

Could I kindly ask you to please:

• create the new issues and link to them from here
• link to other finalized fixes (tickets) in BioLink

Thank you! :)

from monarch-ui.

In answer to @kshefchek I think that currently we do not distinguish between the isoforms of a gene/protein, and so there is a 1:1 mapping from protein to gene. For right now, it would be fine to have uniprot's forward to the appropriate gene page. With new technologies such as long-read sequencing, it is likely that the community will try to do a better job in the future, and so we should try not to bake the idea of there being a 1:1 correspondence between protein and gene deeply into the logic of Dipper or of the ui.

from monarch-ui.

Function associations are being displayed - though some oddities remain. Those oddities have been listed in #167. Closing here.

from monarch-ui.

I am again getting the error we thought was fixed here, now for worm. For example https://beta.monarchinitiative.org/gene/WormBase:WBGene00011095#function

Sadly, @kshefchek and @deepakunni3 - I have to re-open.

from monarch-ui.

I think this is line with what we found the other day, the counts come from the monarch go ingest, and the table comes from biolink via one of the GO solr cores,
https://monarchinitiative.org/gene/WormBase:WBGene00011095#functions

from monarch-ui.

Whats the solution here? Noticing the count discrepancies as well.

from monarch-ui.

can we replace with monarch-initiative/biolink-api#318?

from monarch-ui.

Even #167 is related to this issue.

from monarch-ui.

can we replace with biolink/biolink-api#318?

Ok. Closing in favour of monarch-initiative/biolink-api#318

Leaving open #167

from monarch-ui.