Diseases which are also Genes

[rgb4268]

I have found some instances of diseases, e.g. MONDO:0020119 which have a child which is actually a gene - in this case ZC4H2. There are a few more.

[nicolevasilevsky]

@kshefchek did you say this is actually an issue with Monarch, not Mondo?

[kshefchek]

I suspect it's an integration bug with merging Ids from multiple sources, I can take a closer look

[kshefchek]

There are a few more.

@rgb4268 could you list these here? I can only find this one case.

[TomConlin]

We do have the data structures on hand to "trust but verify" HGNC's assertion that their OMIM dbxref is in fact to what we also believe is a gene

[kshefchek]

false alarm on HGNC, still trying to trace this one

[kshefchek]

closer look, this looks like an NCBIGene issue: https://www.ncbi.nlm.nih.gov/gene/55906 owl:equivalentClass http://omim.org/entry/309605

and this is equivalent to MONDO:0010666

looks like we need to extend the omim type checking down to this line: https://github.com/monarch-initiative/dipper/blob/master/dipper/sources/NCBIGene.py#L446-L466

[TomConlin]

A bit more context omim:309605 has been subsumed into
omim:314580 which is for WIEACKER-WOLFF SYNDROME but the original may have had a different role when it was assigned to NCBIGene:55906

Dipper does handle the replacement & classifies it as a disease

Also we no longer create an equivalence class as it no longer falls through with continue

[pnrobinson]

Note also this entry https://omim.org/entry/301041 I think it is confusing to separate the diseases into XLR and XLD as is being done here

[TomConlin]

close w/ #964