cmungall
commented
7 years ago Hi Michael,
Yes, you are correct that the root of this can be traced to the fact that Orphanet makes two xrefs with the 'E' category, one to the generic form, the other to the AR form.
note that we take a probabilistic approach. We use the orphanet e/narrow/broad assignments to boost prior probability. Ideally the best globally consistent interpretation of all xrefs is found.
in this case it's not clear why the algorithm didn't make the right choice. not only is there other strong evidence pointing to Orphanet:261 as being equiv, it's globally better. Am investogating...
On 7 Dec 2016, at 9:10, Michael Williamson wrote:
Monarch seems to consider Orphanet:98855 equivalent to MESH:D020389. However, the Orphanet term is "Autosomal recessive Emery-Dreifuss muscular dystrophy", while the MeSH term is "Muscular Dystrophy, Emery-Dreifuss". It seems to me that the correct mapping is that MeSH term to Orphanet:261, which is "Emery-Dreifuss muscular dystrophy".
I wonder if this is a specific instance of a more general problem? Specifically, Orphanet explicitly maps both of those Orphanet terms to that MeSH term (as being exactly equivalent). However, I've found that means that Orphanet term is equivalent to the MeSH term or one of its entries, which may be narrower than the main term. In this case, D020389 has "Autosomal Recessive Emery-Dreifuss Muscular Dystrophy" as a narrower entry. In other words, perhaps the MeSH-Orphanet mappings cannot be assumed to be exact mappings during the ontology merging? (All assuming that you're actually using those mappings, and that's the cause of the incorrect equivalence, and that the equivalence is actually incorrect.)
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/monarch-initiative/monarch-disease-ontology/issues/164
mwilliamson-healx
nicolevasilevsky
Monarch seems to consider Orphanet:98855 equivalent to MESH:D020389. However, the Orphanet term is "Autosomal recessive Emery-Dreifuss muscular dystrophy", while the MeSH term is "Muscular Dystrophy, Emery-Dreifuss". It seems to me that the correct mapping is that MeSH term to Orphanet:261, which is "Emery-Dreifuss muscular dystrophy".
I wonder if this is a specific instance of a more general problem? Specifically, Orphanet explicitly maps both of those Orphanet terms to that MeSH term (as being exactly equivalent). However, I've found that means that Orphanet term is equivalent to the MeSH term or one of its entries, which may be narrower than the main term. In this case, D020389 has "Autosomal Recessive Emery-Dreifuss Muscular Dystrophy" as a narrower entry. In other words, perhaps the MeSH-Orphanet mappings cannot be assumed to be exact mappings during the ontology merging? (All assuming that you're actually using those mappings, and that's the cause of the incorrect equivalence, and that the equivalence is actually incorrect.)