Sync NCBIGene updates - Githubissues

monarch-initiative / mondo

Mondo Disease Ontology

http://obofoundry.org/ontology/mondo

Creative Commons Attribution 4.0 International

237 stars 53 forks source link

Sync NCBIGene updates #8415

Open katiermullen opened 3 days ago

katiermullen commented 3 days ago

While adding gene-related non-human animal inherited diseases from OMIA to Mondo, I noticed that sometimes the NCBI gene ID is updated.

A couple of examples:

OMIA:002206-9615 gene was updated from https://www.ncbi.nlm.nih.gov/gene/610421 to https://www.ncbi.nlm.nih.gov/gene/100855746
OMIA:001485-9615 gene was updated from https://www.ncbi.nlm.nih.gov/gene/487827 to https://www.ncbi.nlm.nih.gov/gene/100855930

I noticed this from a manual review of OMIA term to NCBI gene relations. Is there an automated way that NCBI gene IDs are updated in Monarch? I am asking for future me that will want to ensure that NCBI gene IDs associated with non-human animal genetic diseases in Mondo are up-to-date.

Note: I have informed OMIA of these findings so they could update their database.

Thank you!!!

matentzn commented 3 days ago

I think I need this explained again.. This sounds very wrong. Where do the NCBI gene ID get updated? in OMIA you mean?

There is no automated way NCBI gene ids are updated in Mondo!

sabrinatoro commented 3 days ago

@matentzn I "think" it is similar to when an ontology term is merged and/or replaced by another term. In Mondo, the terms that are merged are obsoleted and have the annotation "replaced by" with another Mondo ID. This way, anyone who is using these terms can update the ID they used.

I don't know that there is anything similar to "replaced by" information in NCBIgene, but if this information exists, we should use it to update the information in Mondo and in Monarch.

matentzn commented 3 days ago

That would be the question to @kevinschaper! @kevinschaper is there a way to include "replaced by" relations in the ncbigene ingest for this purpose?

I am also wondering about the same question in Monarch KG:

shouldn't ncbigene->ncbigene and hgnc->hgnc replaced by links be included in the gene mappings SSSOM? So that replaced gene ids are rewired and thereby integrated?

kevinschaper commented 2 days ago

I know that ZFIN’s file for that is here https://zfin.org/downloads/zdb_history.txt, I’ll look into where we can find more. You’re definitely right about making SSSOM and rewiring based on updated IDs