Help extracting ZFIN phenotype data

A request for data from Jim Balhoff (our first analysis user?)

Jim, working with Wyatt Clark, needs a semantic similarity measure for all ZFIN gene pairs based on phenotypic similarity. Wyatt has his own code for semsim, but it assumes a single ontology of precoordinated terms.

Jim suggested doing this via the Uberon parent of every ZFA term in the ZFIN gene to phenotype tables (essentially ignoring PATO):

" So I'm wondering if you have any ideas about the simplest way to collapse things to the nearest E and Q, where the E is the closest Uberon superclass, given that the data actually reference ZFA or HPO. It's surprisingly hard to extract a report like this out of a triplestore.

Alternatively, could you point me to anything ready-made at the LAMHDI site?"

Actually, I think what might be best for this analysis is the ZFIN gene to UberPheno table available from the HPO site. Currently we don't ingest this. I had envisioned ingesting a table that was a mapping from a ZFIN phenotype 6-tuple [1] to the HPO CrossSpeciesPheno file and using this to derived gene2crossSpeciesPheno from our own ingest of the ZFIN table. But I guess it's simpler to just ingest the table directly from HPO[2,3]. And for Jim's purposes to bypass us as the middleperson...

I think we will want to do the extra ingest ourselves soon.

Refs:

• [1] https://docs.google.com/document/d/1Vbokc9aFHR4awNE6DrrLtgpE6axeTS4VEfxqDHsWyPQ/edit#heading=h.ymyapzu4c40w
• [2] http://f1000research.com/articles/2-30/v1
• [3] http://compbio.charite.de/hudson/job/hpo.ontology.uberpheno/lastStableBuild/artifact/data/uberphenoOut/