We recently obsoleted DOID:0060963, which was created as a duplicate of dystonia 5 (DOID:0090043), and renamed dystonia 5 to match OMIM's entry for it at MIM:128230. DOID:0090043 also has both autosomal dominant and autosomal recessive inheritance axioms.

The situation with this disease is a bit more complicated than I previously realized and I suggest we revise it in the DO.

OMIM has relevant information in 3 entries for:

1. dopa-responsive dystonia (aka autosomal dominant Segawa syndrome; https://www.omim.org/entry/128230)
2. BH4-deficient hyperphenylalaninemia B (https://www.omim.org/entry/233910)
3. GCH1 gene (https://www.omim.org/entry/600225)

Aside from the inheritance info in the table, OMIM's entry for 'dopa-responsive dystonia' only contains information about the autosomal dominant version of the disease and the corresponding disease in DO, DOID:0090043, mentions only autosomal dominant inheritance in the text definition. OMIM's entry for 'BH4-deficient hyperphenylalaninemia' has an included entry for 'autosomal recessive dopa-responsive dystonia with or without hyperphenylalaninemia'. More info for each of these phenotypes is listed in the GCH1 gene entry.

Here's what I propose:

1. Update DOID:0090043 to reflect that it refers only to the autosomal dominant form of the disease, including remove the autosomal recessive inheritance axiom, and changing the label to 'autosomal dominant dopa-responsive dystonia'. I suggest the current label 'dopa-responsive dystonia' be added a broad synonym.
2. Add a new disease for 'autosomal recessive dopa-responsive dystonia' as a sibling of DOID:0090043, and add the MIM:233910 as a skos:relatedMatch.
   • NOTE: Using skos:relatedMatch could be a new pattern for INCLUDED entities in OMIM records, if we want to create them as diseases (which I think makes sense in this case).
   • ❓❗@lschriml, @sbello what do you think about the general idea of adding OMIM's INCLUDED entities as diseases and using skos:relatedMatch to link them to the OMIM entries they're included in? I know we recently discussed this. In thinking on it further, it has occurred to me that it's easier for everyone to create more diseases and later obsolete and lump them as necessary, than to create fewer and later have to split them. For the former approach, the existence of term replaced by should make it fairly straightforward to update records annotated to a replaced/lumped term in an automated way, while the latter basically requires human review of records.
3. Create a parent for these two diseases named 'dopa-responsive dystonia'.
   • The downside of creating this disease term, is that it might be used when the more specific autosomal dominant or recessive terms are known. The upside is that if it is ever unknown which inheritance version a disease belongs to it can still be annotated to the parent. I think the upside outweighs the downside. ❓ @lschriml, @sbello do you agree?

Also MONDO has a potentially relevant related issue open (issue 8071), which is geared toward aligning with Orphanet.

cc, @csbjohnson