search

monarch-initiative / mondo

Mondo Disease Ontology
http://obofoundry.org/ontology/mondo
Creative Commons Attribution 4.0 International
224 stars 52 forks source link

GARD: Review "proxy merges" #6429

Open matentzn opened 1 year ago

matentzn commented 1 year ago

During our integration of GARD we encountered 242 GARD ids that map to more than one Mondo ID. This means that, Mondo decided to "lump", while GARD favours a "split". Since we trust the GARD team and their curation efforts, we need to review all 242 cases (121 Mondo IDs), and either:

  1. Perform the split
  2. Convince GARD of the merge

Here is the list, sheet name "mondo_gard_duplicates".

joeflack4 commented 11 months ago

Maybe it's clear to others but when I see "mapped" it is unclear to me. In any case just to clarify, these are all skos:exactMatch in that sheet.

matentzn commented 6 months ago

@twhetzel this could be prioritised at some point - I have talked to Eric from GARD, and he would like to be informed about all proxy merges to be able to make decisions based on it.

I think this is the PR relevant to this issue: https://github.com/monarch-initiative/mondo/pull/6491 (the PR frames the issues as "QC" but the report that is generated with the sparql query is, I think, what the table I linked above should be (it should probably be updated).

matentzn commented 6 months ago

Here is the script that generated the data: https://github.com/monarch-initiative/gard/blob/f95482a9451755c24cad7cdc7cb22fd65558ded3/gard_owl_ingest/mondo_mapping_status.py#L204

I would ask Joe to regenerate this google sheet for review by Sabrina (review only of proxy merges)

The PR I linked above did NOT generate this table!