Feedback is required on the level of support from the SNOMED community for the launch of a project.
| Date created: 15 January 2026 |
|---|
| Action: Feedback requested |
| Status For comment |
| Disposition Open |
| Feedback by Feb. 15th, 2026 |
MF Briefing Note_ v.02Addition of new attribute to link Substance to Organism - Google Docs.pdf (237.8 KB)
Hi Monica,
The Netherlands is very much in favour of this effort. We would be willing to contribute a (part-time) terminologist to work on this; our efforts to translate organisms, antigens and antibodies in the LOINC Ontology should be helpful to us. Some comments:
Still: even unamended, the proposal would lead to a quality improvement that we would support.
Monica, we, in the Allergy-Hypersensitivity CRG have explored this issue in the past wrt the modeling of allergy to animals. The issue we encountered was that the causative agent is not the animal itself but rather a substance (generally proteins with a disposition to act as an allergen) produced by the animal. To address this, we created several new concepts of the form of animal x protein (substance). This solution is less than ideal in that allergens represent many different proteins derived from the same animal and not all proteins derived from that animal are allergenic. In addition, there are many allergies to substances derived from other organisms such as plants, venoms, and animal and plant derived foods potentially requiring many additional protein substances to be created. A better approach imo as supported by this proposal would be to resurrect 90260006 |Allergen (substance)| (without its previous descendants) linking it to an organism using has originating organism. Subtypes of 90260006 |Allergen (substance)| could be added if the specific allergenic protein is known (e.g. 1157148000 |Arachis hypogaea 2 protein (substance)|, a major peanut allergen). Also, supporting the original proposal would be the use of has originating organism in the pharmaceutical domain to model the 709079002 |Allergen extract (product)|hierarchy.
Thanks
The broad notion of formally defining and structuring organism-derived substances in terms of their originating organisms sounds to be a valuable and worthwhile exercise, therefore The UK is interested in being involved in the work outlined in the briefing note. However more information is needed regarding the priority use cases before we know which experts to involve.
The briefing note talks most about a ‘laboratory’ use case, but it is likely there is also an ‘allergy/adverse reaction/toxicity’ use case [including Bruce’s contribution above], based on relatively recent discussions here and here there may even be elements of a ‘herbal medicines’ use case, and the discussion here notes the challenge of handling synthetically engineered substances as opposed to those of natural origin. Each of these use cases may have differing requirements so may need a different solution.
Spelling out each priority use case clearly (which user groups, which activities, which outcomes) will be a helpful precursor to understanding and testing the suitability of the proposal, before any resource is expended in progressing to the steps given in the ‘Proposed solution’ section in this briefing note or the much more detailed ‘project approach’ in the earlier Business case proposal.
Kind regards
Ed
I agree with the approach outlined by @bgoldberg . Though I’d argue the animal x protein (substance) pattern might be too specific for some use cases - a more general animal x substance (substance) would have broader application - though both patterns can co-exist (protein being a specific substance). It’s not always known if a protein caused a (allergic or hypersensitivity) reaction or some other substance.
Somebody with 1306845001|Alpha-gal syndrome| might have a reaction after eating a pork chop. But I don’t think it’s accurate to call that a reaction to |swine protein| - (unless we adopt the broader culinary definition of protein).
But especially with reported reactions (for which we don’t know the specific hypersensitivity process) - I don’t think having concepts with such specificity (in the absence of the general) is as useful.
I’d expect 418626004|Allergy to lobster| to be chosen over Allergy to lobster protein by most end users (patients and clinicians).
The Canadian team is interested in this proposal for the medicines use case (as suggested by @echeetham) - particularly in the representation of substances contained in licensed natural health products. The Canadian team would like the medicines use case to be considered, and support this proposal on this basis.
@mcordell I completely agree with you in terms of improving the model of allergy conditions by making it less specific. At the time, discussions in the CRG including some allergy specialists in addition to myself concluded that the causative agent (allergen) for most inhalant animal allergies was most likely a protein. The example of alpha-gal represents a food allergy caused by a carbohydrate allergen. My proposal is to do away with animal x protein by using a causative agent of simply Allergen as a subtype of Antigen (substance). The Allergen substance hierarchy was inactivated several years ago I think because it was felt that any substance could be an allergen (not true). I believe one could define Allergen substance with a has disposition of allergen (disposition) similar to the way Toxin (substance) has been defined. The causative agent of allergic conditions would thus be defined simply as caused by allergen (substance) and this could then be linked to the organism with has originating organism without the need for creating separate organism specific allergen concepts. As I mentioned subtypes of Allergen (substance) could be created when the specific component allergen is known e.g. alpha-gal, Are h 2 peanut, etc. I am proposing by invoking this model that we eventually retire the Animal x protein concepts we previously created, Allergy to animal X could be used as a PT or perhaps as the FSN as well by including a text definition.
Bruce
The Swedish team supports the proposal to link the Substance hierarchy directly to the Organism hierarchy as proposed by other member countries in the various used-cases above. Allergy is a good example. In laboratory medicine, the Source organism would be the subject of the test by default.
Kind regards,
Erica and Keng Ling @klwallin
The Spanish team supports the proposal to link the Substance hierarchy directly to the Organism hierarchy. In addition, we consider that having this attribute available would be highly valuable for us, particularly for the modelling of concepts representing herbal medicinal products, where explicitly linking substances to their originating organisms would greatly improve consistency.
Kind regards,
Laura Cristino and Marta Pavón
The New Zealand team supports linking the Substance hierarchy directly to the Organism hierarchy.