Some gene variants do exist in SNOMED CT, e.g. NPU component+specification HLA-DQB1 gene(DQB1*03:02) could be mapped to 257118006 |DQB1*0302 (substance)| in:
NPU56087 DNA(spec.)—HLA-DQB1 gene(DQB1*03:02); arb.entitic num.(proc.) = ?
But no corresponding concepts for components in below NPU concepts:
NPU29745 DNA(spec.)—HLA-B27 gene; arb.entitic num.(proc.) = ?
NPU58055 DNA(spec.)—CYP2D6 gene(NG_008376.3:g.6750); seq.var. = ?
Genetic tests out of scope of LOINC ontology for now…
Hi Daniel, do you mean it is Ok to map the NPU component+specification HLA-DQB1 gene (DQB1*03:02), from a gene to a substance hierarchy ? or should we follow the new recommendation and create a concept under cell structure?
First, lets discuss tomorrow.
My understanding is that the first NPU concept refers to a specific variant 03:02 not the gene.
Hi,
I have been doing some corresponding with Jim Case about creating new SNOMED CT concepts for genes. In view that the discussion on where to create gene concepts under the body structure hierarchy is still ongoing, his advice to me was to specifically not to create them in the substance hierarchy. With that, the issue is solved, since I was asked to come up with a solution, my solution or suggestion is that we leave out these NPU codes from the pilot project and replaced them with 3 others that are commonly used in among the Nordic countries. Once this issue of modelling gene concepts have been around and comments and concerns have been collected and summarized and approved by S International, then we would be able to solve all the 1000 or more gene NPU codes. In the meanwhile, I create SNOMED codes for category and taxon. Please voice your opinion regarding the proposal to have replacements instead of the NPU codes for genetic analysis.
Does this still apply to
| NPU56087 | DNA(spec.)—HLA-DQB1 gene(DQB1*03:02); arb.entitic num.(proc.) = ? |
where there is a mapping to the variation, not the gene?
Anyway, I picked three new NPU concepts randomly from the high-usage list (easy ones, but not very interesting…):
| NPU01157 | P—Aluminium; subst.c. = ? µmol/L |
| NPU28309 | B—Haemoglobin; mass c. = ? g/L |
| NPU28315 | Ercs(B)—Haemoglobin; mass c. = ? g/L |
After en discussion with those who used to work in the field av genetic testing, the variation is not a substance but a variation in DNA sequence when expressed in this form, HLA-DQB1 gene(DQB1*03:02).
I will have a look at these new NPU codes and its elements in the meantime.
I received a reply from Jim Case regarding the dateline for review of the proposed modelling for genes, It will be the 31 December 2025. Please help us by circulating the proposal to those working directly or are advisors in the field of human genetics.
When SNOMED International has summarized the comments gathered on the principles for modelling genes, we will have an exciting time incorporating the new concepts into NPU which will have en impact internationally and across other code systems.
I am including this ref to help understand what then lab tests are and what DQB1*03:02
denotes, as an example.
The HLA-DQB1*03:02 allele is an HLA gene variant that is a genetic marker for certain conditions.
Thanks, Daniel, for the articles. I will save them all for a rainy day. Not all the specifications in LabTerm had a defining text, so I have now finished writing one. I think I will need to adjust the definition slightly according to the terminology used in the articles you shared. I will raise the question in the NPU community
“sequence variation” is the kind-of-property. The variant in the specified position of the gene is the observed value or finding.
20205-11-21:
New in the 100 set:
NPU04980 Csf—Albumin; rel.subst.c.(Csf/P) = ?
NPU19661 U—Albumin/Creatininium; mass ratio = ? × 10-3
NPU19680 Pt(U)—Albumin; mass rate(proc.) = ? mg/d
