concept · intermediate · partly written
Ontology mappings, cross-references and SSSOM
A mapping is a claim that a term in one vocabulary matches a term in another; SSSOM adds the kind of match, a confidence and where it came from.
Also called: SSSOM, cross-reference, xref, ontology matching
What a mapping is
A mapping is a triple: a subject term, a relationship (the predicate) and an object term, ideally with who made it and how sure they were. The older cross-reference (an "xref") says only that two terms are related, without saying how. 1
Established
Why cross-references fall short more
The SSSOM authors write that cross-references "are used in vastly different ways in different ontologies" and cite an analysis of about a million of them that judged them "impractical or even impossible to reuse". This atlas did not open that analysis itself. 1
Reported, not yet replicated
SKOS and SSSOM more
SKOS (the W3C vocabulary standard of 2009) has five match relations; exact match is transitive, and close match is deliberately not, to avoid compounding errors when chained. SSSOM is a table with explicit predicate, confidence and provenance, with 38 slots per mapping. The SSSOM paper states that Mondo exports SSSOM tables and names Open Targets, NMDC and Biomappings as users. 21
Reported, not yet replicated
Measured error rates more
In an analysis of BioPortal mappings converted to equivalence axioms, logical conflicts appeared in 11 of 19 ontology pairs, with on average 22% of mappings per pair involved; of the mappings that automatic repair removed, over 40% were erroneous, and error was the cause of removal in over 60%. "Involved in a conflict" does not mean wrong. For OMOP2OBO, expert review of a random 20% sample found 73.9% of the manually derived condition mappings correct (2,000 reviewed) and 70.7% of drug-ingredient mappings (116 reviewed); the authors say the manual mappings were made by a non-clinician and may contain errors. For measurement results, 97.3% of 1,350 were correct. 34
Reported, not yet replicated
Machine-assisted mapping more
Biomappings, which predicts and curates missing identifier mappings, found 570 correct, 10 incorrect and 6 unsure among 586 reviewed lexical predictions for one cell-line pairing. On the hardest disease task in the Bio-ML benchmark, four of the best unsupervised matchers scored F1 of about 0.61 to 0.64 against a reference set the authors call imperfect; a plain edit-distance baseline matched or beat trained systems on one task. 56
Reported, not yet replicated
What mappings buy more
Adding 2,909 MeSH-to-ChEBI mappings made 100,342 ClinicalTrials.gov trials (70.6% of trials with an intervention annotation) linkable to 995 chemicals; that shows reach, not correctness. In five rare diseases in the All of Us data, queries built on OMOP2OBO mappings found 99.3% of the patients that validated manual mappings found, with fewer codes and about a third of the time. 54
Reported, not yet replicated
Not found more
No controlled experiment that seeds known mapping errors into a graph and measures the change in a downstream task was found. Damage from wrong mappings is shown only as logical conflicts and by named examples. The SSSOM paper gives no error rate. 13
Reported, not yet replicated
Plain answer
Why it matters
Still open here
- What is the real error rate of cross-references and SSSOM mappings across the OBO ontologies?
- Does an error-injection study exist that measures the downstream cost of wrong mappings?
Not yet written: Equation, Real numbers.
Linked from
Sources
- Matentzoglu N, Balhoff JP, Bello SM, Bizon C, et al.. A Simple Standard for Sharing Ontological Mappings (SSSOM). Database : the journal of biological databases and curation (2022). paper; opened 2026-11-18 all records citing this
- SKOS Simple Knowledge Organization System Reference.. report, not peer reviewed; opened 2026-11-18 all records citing this
- Towards annotating potential incoherences in BioPortal mappings.. paper; opened 2026-11-18 all records citing this
- Tiffany J. Callahan, Adrianne L. Stefanski, Jordan M. Wyrwa, Chenjie Zeng, et al.. Ontologizing Health Systems Data at Scale: Making Translational Discovery a Reality. arXiv (2022). preprint, not peer reviewed; opened 2026-11-18 all records citing this
- Hoyt CT, Hoyt AL, et al.. Prediction and curation of missing biomedical identifier mappings with Biomappings. Bioinformatics (Oxford, England) (2023). paper; opened 2026-11-18 all records citing this
- Yuan He, Jiaoyan Chen, Hang Dong, Ernesto Jiménez-Ruiz, et al.. Machine Learning-Friendly Biomedical Datasets for Equivalence and Subsumption Ontology Matching. arXiv (2022). preprint, not peer reviewed; opened 2026-11-18 all records citing this
Where it sits on the map
History of this record (6)
This record has been changed 6 times, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.
- Status
- partly written
- Last updated
- 2026-11-19
- Record
- content/concepts/ontology-mapping-and-sssom.json