The US and Australia have put forward a business case proposal (see attached) for creation of a new attribute for the Substance hierarchy which can be used to link a biological substance with the organism it is sourced from. E.g: Taipan venom (substance): IS A Venom (substance) , has originating organism = Oxyuranus scutellatus (organism). The topic was taken to the DEUSG in October and received support from the group. https://conf.spaces.snomed.org/wiki/spaces/USRG/pages/156598382/2025-10-20+DEUSG+Antwerp
Comments: This would help to clarify how organisms and their parts are used as ingredients in medications. The range of “Has active ingredient” remains unchanged to substances, and the organisms would be linked using the new relationship. The substance used as a medication ingredient can be the a protein, an extract, etc of the corresponding organism. The relationship between the two hierarchies would help to resolve any apparent duplication between these hierarchies.
The CMAG is being approached to see if there is support for this work item. Please advise if your country supports this work and if not, please advise of any concerns you may have.
We don’t have any specific use cases for this connection but we understand the value and potential for decision support etc. that the proposed attribute would allow. We support the work on a theoretical level, even if we don’t have a pressing need for it nationally at the moment.
Similar topic was presented by Stan Huff to the LOINC laboratory committee meeting last month in which he was purposing to remove “Antigen” information from different test in LOINC in order to address the lack of an attribute relationship between the substance and organism hierarchy. The general consensus from the laboratory community was that the identification of antigen, antibody, RNA, DNA, etc. type substances is important to the laboratory community so the general consensus was not to change LOINC but to add the attribute to SNOMED CT to link the two hierarchies. The LOINC ontology use-case aligns with the use-cases presented by both the US and Australia. Can this additional information regarding the LOINC ontology be taken into consideration for this topic?
What does it mean to say that a substance “has source organism”
Is it:
a. the biological identity the substance is based on
or
b. the physical source it was made from
From Business Case Proposal for CMAG
• Pork insulin (substance): IS A Insulin (substance), has originating organism = Sus scrofa scrofa (organism)
• RNA of Chikungunya virus (substance): IS A Ribonucleic acid (substance), has originating organism = Chikungunya virus (organism)
Insulin that has a porcine (pig) or bovine (cow) amino-acid sequence does not necessarily come from pigs or cows anymore. This type of insulin can be made using bacteria or yeast that have been engineered to produce the insulin sequence, or chemically synthesised and then processed.
In these situations:
• the insulin still has a pig or cow insulin sequence
• the organism that actually produces it is not a pig or a cow
• no animal tissue is involved at all.
RNA of Chikungunya virus refers to genetic material that matches the virus. However, the RNA used is often made in a lab or copied many times during testing, rather than taken directly from the virus itself.
These are very valid questions. When we wrote the proposal, I think we are suggesting “Has source organism” as a possible attribute name to open the discussion, not as a final recommendation. We are open to a better name for the attribute that covers the “derivative of ” use case that we are going to encounter with genetically modified or laboratory produced products. This is a slippery slope because there may need to be more than one attribute to define a necessarily true condition between substance and organism hierarchy. We may also need to reclassify some substances to represent “synthetic” variants to better address the issue raised regarding insulin produced by E. Coli or Saccharomyces cerevisiae. My initial thought would be that a “synthetic porcine-like insulin derived from Saccharomyces cerevisiae (substance)” might be modeled with a “Derived from” attribute pointing to Saccharomyces cerevisiae, but then this begs the question of whether this level of information is available from the manufacturers?
This is a good question and not something we’d really thought of (at least I hadn’t).
As John said, I don’t think any of us are committed to a particular name for the attribute, just that there should be some association.
I think for over 95% of use cases the “biological identity” and “physical source” are one and the same. My inclination is to not over-complicate this and take the concepts at face value.
“Pork insulin (substance)” - is insulin, from a pig.
From my brief research, it looks like this is still true. The recombinant sources are identified as such - “recombinant”. Often(Usually?) there’s no qualifying information (human/porcine) on recombinant insulin since that’s now the predominant manufacturing process. Animals were the predominant source before we could manufacture the human analogs.
I think John’s description “Synthetic porcine-like insulin derived from Saccharomyces cerevisiae (substance)” is the most accurate term if we wanted that level of detail. With “porcine-like” suggesting it’s a sibling, not subtype of porcine insulin.
“RNA of Chikungunya virus” - is RNA from Chikungunya virus.
I’m not sure of a use case where it’s useful to differentiate between the “original” RNA and amplified material. Amplification is implicit in the process, just as subculturing is for bacteria.
The use cases documented in the proposal are satisfied with this approach:
“Chikungunya virus nucleic acid assay (procedure)” and “Chikungunya fever (disorder)” could have traversable relationships between them. Currently there’s nothing.
“Pork insulin (substance)” - maybe not as useful, except perhaps for religious purposes (or maybe allergies)? But the property still allows concepts to be sufficiently defined and classified, and unambiguous.
Consider: 303318006|Funnel web spider venom (substance)| - which spider is this referring too? North American’s might say Hololena curta, Chinese might say it’s Macrothele sp. and Australian’s will say it’s a member of the Atracidae family.
As an aside: We (AU) created 422831006|Gymnorhina tibicen (organism)| when we realised 43428003|Magpie (organism)| is not this animal responsible for over 500 reported injuries last year.