MONDO:0034146 and MONDO:0032733 are the same disease (GLS repeat-expansion glutaminase deficiency) — merge

[cmungall]

The problem

Mondo has two terms for one disease:

Term	Label	Provenance
MONDO:0032733	global developmental delay, progressive ataxia, and elevated glutamine	OMIM:618412, MEDGEN:1680160, UMLS:C5193080
MONDO:0034146	spastic ataxia-dysarthria due to glutaminase deficiency	Orphanet:557056, GARD:0022247, MEDGEN:1814477, UMLS:C5681336

These are the same entity — the autosomal recessive GLS repeat-expansion disorder reported by van Kuilenburg et al. Orphanet itself cross-references Orphanet:557056 to OMIM:618412; Mondo never picked that up, because the two terms entered from separate ingests a month apart in 2020 and were never reconciled.

Proposed fix: obsolete MONDO:0034146 with replaced_by: MONDO:0032733, migrating its synonyms, xrefs, subsets and logical axioms onto the surviving term.

This is Phil Richmond's finding — he worked the whole thing out in a slide deck walking through the GLS disease series in OLS and OMIM, spotted that MONDO:0034146 had no OMIM xref, and traced Orphanet:557056 back to OMIM:618412. Everything below is confirmation of his analysis. Thanks Phil.

Detail

GLS currently has four disease terms; there are three diseases

GLS variants cause three recognised conditions:

1. CASGID (MONDO:0032685, OMIM:618339) — infantile cataract, skin abnormalities, glutamate excess and impaired intellectual development; a gain-of-function variant, elevated glutamate:glutamine ratio.
2. DEE71 (MONDO:0032678, OMIM:618328) — neonatal-lethal developmental and epileptic encephalopathy; biallelic loss of function.
3. GDPAG (MONDO:0032733, OMIM:618412) — global developmental delay, progressive ataxia, elevated glutamine; biallelic, driven by a GCA repeat expansion in the GLS 5'UTR.

MONDO:0034146 is a fourth term for the third disease.

Evidence that MONDO:0034146 and MONDO:0032733 are the same

Orphanet asserts the equivalence. ORDO's record for Orphanet:557056 carries database_cross_reference: [ICD-10:E88.8, ICD-11:5C50.Y, MONDO:0034146, OMIM:618412, UMLS:C5681336]. The OMIM link is explicit — no inference required. Mondo's MONDO:0034146 simply lacks the OMIM xref that its own source vocabulary provides.

One primary source, three patients. Both terms trace to van Kuilenburg et al., NEJM 2019 (PMID:30970188), which reports the GCA repeat expansion in three unrelated probands. The paper's own summary — "an early-onset delay in overall development, progressive ataxia, and elevated levels of glutamine" — is the OMIM:618412 title verbatim. There is no second cohort for MONDO:0034146 to describe.

The definitions describe one disease. MEDGEN:1680160 (OMIM side): early-childhood delay of gross and fine motor skills and speech, ataxia by mid-to-late childhood necessitating walker or wheelchair, plasma glutamine persistently elevated ~2.5-fold, expanded GCA repeat in the GLS 5'UTR. MEDGEN:1814477 (Orphanet side): childhood-onset global developmental delay, progressive spastic ataxia leading to loss of independent ambulation, elevated plasma glutamine, cerebellar atrophy. Same onset, same course, same biomarker, same gene.

OMIM leaves no room for a fourth entity. The CASGID entry (OMIM:618339) states that "GLS loss of function is implicated in developmental and epileptic encephalopathy-71 (DEE71; 618328) and a syndrome of global developmental delay and progressive ataxia (GDPAG; 618412)." Two LoF diseases, not three.

The two MedGen/UMLS pairs are the usual duplicate signature: one concept minted from OMIM, one from Orphanet, for a single disease.

How it happened

Term	Introduced	Commit	First release
MONDO:0034146	2020-03-26	3308f2dd7 — "ORDO NTs #1339", bulk Orphanet ingest	v2020-04-05
MONDO:0032733	2020-04-28	548116e0b / f2c173ec6 — "omim-import"	v2020-04-29
MONDO:0032678	2020-04-28	same OMIM import	v2020-04-29
MONDO:0032685	2020-04-28	same OMIM import	v2020-04-29
MONDO:0600001	2021-05-06	61772ad0a — manual, kallia-p	v2021-06-01

The Orphanet term landed first. Three weeks later the OMIM ingest minted a fresh term for OMIM:618412 rather than attaching the xref to the term already in the file, and the duplicate has been sitting there since.

This also explains the hierarchy asymmetry Phil noticed. When glutaminase deficiency (MONDO:0600001) was created in May 2021, MONDO:0034146 and MONDO:0032678 were nested under it, but MONDO:0032733 was missed and still hangs off bare hereditary disease with no definition. Its thin hierarchy isn't a modelling decision — it's an un-curated ingest artifact.

One thing not to carry over

ORDO annotates Orphanet:557056 as has_inheritance: Autosomal dominant. That is wrong: all three families in PMID:30970188 are biallelic — two compound heterozygous, one homozygous for the expansion. Mondo already has this right, placing MONDO:0034146 under autosomal recessive spastic ataxia (MONDO:0017847). The AR placement should survive the merge; the ORDO inheritance annotation should not be propagated.

Proposed changes

Obsolete MONDO:0034146 as a merge into MONDO:0032733:

• is_obsolete: true, replaced_by: MONDO:0032733, property_value: IAO:0000231 MONDO:TermsMerged
• strip all logical axioms (nothing in Mondo currently references MONDO:0034146, so no rewiring is needed)

Migrate onto MONDO:0032733:

• the Orphanet label as an exact synonym, and GARD:0022247, MEDGEN:1814477, Orphanet:557056, UMLS:C5681336
• subsets gard_rare, nord_rare, ordo_disorder, orphanet, orphanet_rare, rare
• is_a MONDO:0600001 (glutaminase deficiency), MONDO:0017847 (autosomal recessive spastic ataxia), MONDO:0019052 (inborn errors of metabolism)
• a definition, which the term currently lacks, cited to PMID:30970188

Leaving GLS with three disease terms under glutaminase deficiency and its GoF sibling.