search

EBISPOT / efo

Github repo for the Experimental Factor Ontology (EFO)
https://www.ebi.ac.uk/efo/
54 stars 14 forks source link

Reviewing equivalent to axioms for disease terms #295

Open zoependlington opened 5 years ago

zoependlington commented 5 years ago

Prompted by a discussion regarding nervous system disease and central nervous system disease:

I decided to look at the top level disease terms to see what the equivalence was looking like. Many of them need editing or axioms adding (blue), there are a few terms that need to be reclassified (yellow), a few I couldn't find a quick solution to and may require more investigation (pink) and a handful that are okay (green).

I've added all the suggested equivalent to axioms in the Equivalent column, although it would be nice to have a blanket way of expressing these axioms going forward.

Spreadsheet can be found here: https://docs.google.com/spreadsheets/d/1p_NTmbNMR-UY9ZWXsI7OxxGmGWpAG0u_MU3hyL2xAYs/edit?usp=sharing

paolaroncaglia commented 5 years ago

Hi @zoependlington , I went through the spreadsheet and added comments in the "Equivalent" column. My comments are all preceded by "[Paola]" to avoid confusion. If anything is unclear, please let me know. Thanks!

paolaroncaglia commented 5 years ago

At our EFO meeting today, we resolved that we should adopt (more) consistently the design pattern recorded here: https://github.com/simonjupp/efo3-creator/tree/master/designpatterns. In particular, the one for disease location is the following - and we should make sure that the brackets are the same across all equivalence axioms of the same type:

equivalentTo: text: "%s and (has_disease_location some (%s or ('part of' some %s)))" vars:

Also, it will help when a fresh import of UBERON is pulled into EFO, so perhaps wait for this ticket to be done: https://github.com/EBISPOT/efo/issues/310

paolaroncaglia commented 5 years ago

Updating this ticket to add that David and Nico found more axioms that need fixing, see point 3 in https://github.com/EBISPOT/efo/issues/372#issuecomment-470908980. Zoe and I suspect that there will be more. We could probably pull them all out with a query. I'll move this ticket nearer the top of the backlog.