search

monarch-initiative / monarch-mapping-commons

Building a fully exectuable workflow for boomer
Creative Commons Zero v1.0 Universal
4 stars 3 forks source link

WIP: Add LOINC PART mappings #35

Open matentzn opened 1 year ago

matentzn commented 1 year ago

@drseb this is work in progress, I will give you some rounds of feedback on these.

Dont change the skos:exactMatches yet.

drseb commented 9 months ago

@matentzn please give another round of feedback and help with the open issues. thanks a lot

drseb commented 8 months ago

I hope it looks better now

matentzn commented 8 months ago

@drseb can you confirm:

You personally curated all of these mapping sets manually? Can you write two sentences here (in a comment on the PR) on how you created the mapping sets exactly (methodology), and what mapping rules you applied (using exact in case where units were mentioned, etc).

drseb commented 8 months ago

All mappings except loinc2chebi_1.sssom.tsv were created using a semi-manual approach, in which software suggested mappings based on string comparisons, but a human would always review and possibly refine the mapping. loinc2chebi_1.sssom.tsv was obtained from compLoinc (Tim etc.)

matentzn commented 8 months ago

Can you describe in a bit more detail the mapping rules you used, for example:

In cases where the LOINC part reflected a measurement like protein^3rd tube we used skos:exactMatch to map to PR:protein to reflect the fact that the entity described by the LOINC part is identical to the PR ontology concept.

but a human would always review and possibly refine the mapping.

So so the statement would be correct that all mappings aside from Tims were manually reviewed by you right?