Look at use of ICHPT for translating between phenotype ontologies

[Relequestual]

ICHPT mappings have not yet been finalised or officially released.

We have been advised by Ada to wait for the paper and official website release (digestable format) of the mappings before seeking to implement usage.

[allasm]

Potentially related to #52

[buske]

Does this mean that we should remove ICHPT from the upcoming version of the API we're trying to release?

[fschiettecatte]

I am ok with removing it for now given that we are waiting on Peter for his input/blessing. I am not going to remove support from GeneMatcher since it is already in there, so we will accept ICHPT terms based on Ada's mappings, but we will only emit HPO.

[buske]

+1

It's a backwards-compatible change to add it back in later, so I think we should remove it for now.

[Relequestual]

ICHPT has since been released!

Is there any system in MME that this would be benificial for? ICHPT allows for mapping, but HPO is more descriptive! There's no reason a system can't get a copy of ICHPT (now) and then translate to their phenotype ontology of choosing that it supports.

I think it would actually be detrimental to use ICHPT as the across the wire format, as information would be lost going from HPO to ICHPT and back to HPO.

[fschiettecatte]

It makes most sense to use HPO only for MME going forward. I don't think anyone has expressed an interest or need in using anything else. ICHPT has value but not for MME.

[Relequestual]

Voting to close this issue.

[fschiettecatte]

+1

[buske]

+1 Let's close this until groups have clear use cases that require it.

[buske]

Closing due to 3 votes.