mondo icon indicating copy to clipboard operation
mondo copied to clipboard

Published 20 hours ago • verified publisher icon monarch-initiative

scope of processing from OMIM

Open maglott opened this issue 3 years ago • 10 comments

Submitters to ClinVar and GTR often provide the name of a disorder according to the value they see in genemap2.txt. In some cases, these names do not match any preferred , alternate, or included term in a record from OMIM itself. Are there any plans to integrate the unique terms from genemap into Mondo? NCBI often retains those terms as distinct entities, e.g. https://www.ncbi.nlm.nih.gov/medgen/CN301185

vs. https://www.omim.org/entry/135400 /MONDO:0007610/C1851120

Any opinion about whether or not that is appropriate? Thanks

maglott avatar Feb 21 '22 15:02 maglott

@ahamosh @nicolevasilevsky Would be very interested to hear your thoughts this topic.

kanems avatar Feb 22 '22 16:02 kanems

@ahamosh said that genmap2 comes from OMIM and the disease names should map. @maglott could you please share the cases where you see mismatches with OMIM?

nicolevasilevsky avatar Mar 08 '22 21:03 nicolevasilevsky

I am not saying there is a mismatch; I am saying there are terms unique to genemap2 and not in Mondo. Let me do some analyses and return a list. Is genemap2 included in the slurps that Mondo does? (My first use of slurp). And please remind me of the path to the documentation.

maglott avatar Mar 09 '22 12:03 maglott

I think I've worked out query that provides some signal. I'll send a list next, with the following caveats:

  • NCBI has not processed the March release yet

  • NCBI needs to understand the use of 'has_narrow_synonym' because at present we are not exacting these terms for evaluation. Thus this file over-reports gene_map_names.xlsx

  • {} and ? were stripped from the values in the genemap2.txt This is a report of Mondo IDs with skos exact matches to a MIM number, for which the name in genemap2.txt does not have an exact string match to any name we extract from Mondo (case insensitive).

maglott avatar Mar 09 '22 15:03 maglott

Megan and I have been reviewing some of the terms unique to genemap2. One example is in https://www.omim.org/entry/158350. MONDO:0019002. genemap2.txt uses 'Lhermitte-Duclos syndrome' not Lhermitte-Duclos disease. It would be great if Mondo provided the answer whether or not Lhermitte-Duclos syndrome = Lhermitte-Duclos disease. Right now MedGen has them as distinct, not having raised this specific question yet.

maglott avatar Mar 14 '22 12:03 maglott

Or, for https://www.omim.org/entry/613703, is 'Microphthalmia with coloboma 6' to be considered equivalent to MICROPHTHALMIA, ISOLATED, WITH COLOBOMA 6? Submitters to the NIH Genetic Testing Registry have used the value from genemap2.txt. Last specific example. Hoping for a policy.

maglott avatar Mar 14 '22 12:03 maglott

Thanks @maglott. I'll review this with @matentzn asap.

nicolevasilevsky avatar Mar 18 '22 16:03 nicolevasilevsky

@ahamosh - could you clarify why there are different names in the genemap2 file which are not included the usual OMIM metadata on the website (the file we pull from is called mimtitles.txt) ? We'd like to use the authoritative source for pulling in OMIM synonyms.

@maglott we can add the synonyms for you from the genemap2 file that you shared.

nicolevasilevsky avatar Mar 25 '22 16:03 nicolevasilevsky

Note to self, this is the ROBOT template: https://docs.google.com/spreadsheets/d/1GMXDtq2vOLjOVyFYSrquFpxVOu6dJfBCyHzQRGcD6YM/edit#gid=688099271

nicolevasilevsky avatar Mar 27 '22 00:03 nicolevasilevsky

see related ticket: https://github.com/monarch-initiative/mondo/issues/4939

nicolevasilevsky avatar Aug 26 '22 18:08 nicolevasilevsky