Naming a dismech entry

The rules are recorded in design decisions §18 (docs/explanation/design-decisions.md). This skill is the working procedure.

The two rules

1. The name has the same scope as the entry. Ask: could a patient have the condition this name describes without the entry's defining cause? If yes, the name is too broad. An entry bound to a single-gene term but named after a phenotype (Cryptorchidism, Trichotillomania, Normal Pressure Hydrocephalus) fails this test, because each of those names also covers patients with other genes or with no genetic cause at all. Whether the fix is a rename or a rebind depends on what the entry is actually about (see the worked examples below).

2. A single-gene Mendelian entry is named <GENE>-Related <Phenotype> when that name is established. "Established" means at least one of:

   • the MONDO label, or a MONDO EXACT synonym, of the bound term
   • a ClinGen-preferred label (a MONDO synonym that MONDO attributes to ClinGen)
   • a GeneReviews chapter title for the same scope (cache/bookshelf/genereviews.csv)
   • the current OMIM title
   • recurring use in recent primary literature, cited in the entry's notes (add it to synonyms, too)

   This follows ClinGen's dyadic nomenclature (PMID:33417889, PMID:39241757). For a mechanism resource the gene is the root of the pathograph, so putting it in the name tells a reader where the graph starts.

The MONDO label is an input, not the default

The disease_term binding fixes what disease the entry is. The name is dismech's own. A stub's label is the MONDO label, and the stub filename and claim issue title are built from it, so it is easy to carry it forward unexamined. Treat it as one candidate among those above.

MONDO labels can be broader than their own class, out of date with OMIM, or miscased. Some known cases are listed in monarch-initiative/mondo#10836. If you choose a different name because of a problem with the MONDO label, say so in the PR and either add the term to that issue or open a new MONDO issue (the mapping-analysis skill covers upstream corrections).

Procedure

1. Is the entry single-gene? Check the stub's genes: (or the bound MONDO term's has_material_basis_in_germline_mutation_in) and the entry's causative genetic[] rows. If it is not single-gene, only rule 1 applies.
2. Apply the scope test to the current label. If it fails, the name must change. Then decide which side is wrong: if the entry is really about the broad condition, the fix is a broader disease_term and a fuller genetic section, not a new name.
3. Look for an established gene-related name, in the order listed above. Prefer a candidate whose phenotype words match the entry's scope.
4. Reject umbrella names for a single entry. POLG-Related Disorders names a spectrum; Alpers-Huttenlocher syndrome is one point on it. Renaming the entry to the umbrella makes it broader, which breaks rule 1. An umbrella is a candidate for a Grouping or a parent entry with has_subtypes.
5. Check that the phenotype half keeps the scope. A gene-related name with a different phenotype word is often a different lumping. GeneReviews' ATM-Related Cancer Predisposition covers heterozygous carriers, not ataxia-telangiectasia.
6. Keep the old names. The MONDO label, eponyms and the OMIM title go into synonyms, so search and recognition still work.
7. If nothing is established, keep a clear existing name. Do not coin a gene-related name. An eponym or a numbered series name (Spastic Paraplegia 81) is specific, even if it is opaque. Prefer a dyad when one is established, but an opaque specific name is a smaller problem than a broad one.

Form

• Title case, as for every entry: ACTA1-Related Nemaline Myopathy. The gene symbol keeps HGNC case, and Related is capitalised. The corpus mixes -related and -Related; new entries use -Related.
• No parenthetical qualifiers, as with subtype names. Put long forms in synonyms.
• Check casing against the source rather than copying it. MONDO has labels with a capitalised common word partway through.

Worked examples

Situation	Outcome
Brachyphalangy-Polydactyly-Tibial Aplasia Syndrome, bound to a term whose MONDO label is already HMGB1-related ...	Use the MONDO label's gene-related form
Alpers-Huttenlocher Syndrome, where the only gene-related title is POLG-Related Disorders	Keep the eponym; the umbrella belongs on a grouping
Cystic Fibrosis (CFTR)	Keep. The name is already specific to one gene
Cryptorchidism, bound to the INSL3 term	Scope test fails. Decide whether the entry is about INSL3 (rename) or about cryptorchidism generally (rebind)
Protein S Deficiency, covering PROS1-driven hereditary deficiency (AD and AR forms)	The name drops "hereditary", so it also covers acquired deficiency. Use Hereditary Protein S Deficiency; do not narrow it to the AD-only term, because the entry also covers the AR form

Renaming an existing entry

Changing name on its own is safe. Changing the file slug moves every cross-file reference (File:pathophysiology#..., grouping members, comorbidity entries) and needs a history/ record with superseded_by. Until the project decides how slugs follow renames (dismech#15894), change name and synonyms and leave the filename alone.

Reviewing

In a PR that adds or renames an entry, ask:

• Is the name as specific as the entry?
• If the PR mentions a more specific name used in the literature, has it been considered?
• Are the MONDO label and other prior names kept in synonyms?