Open monicacecilia opened 5 years ago
Can we flesh out the requirements for this? Are we talking the breakdown of species that are models of Marfan syndrome? ie some view of the data when you click the filter button here - https://monarchinitiative.org/disease/MONDO%3A0007947#models
Or do we want a way to filter diseases by species that model that disease (in autocomplete or some dynamic search view)
The page you are refering to is good (although the sorting is not perfect in that human and mouse are mixed). The problem page in my opinion is the landing page after a search https://monarchinitiative.org/search/marfan For both pages, it might be better to return the results like human 20 mouse 32 zebrafish 3 etc and then go to the details -- (question for Julie). At least for me there is too much information here and it might be nicer to return the information in a structured way.
The problem can't be solved on the front end alone; the data are not yet there. First and foremost, MONDO must distinguish human and animal diseases.
You can tell what is a non-human disease. This is sufficient for a useful filter.
@cmungall I'm not sure what you're suggesting. Are you suggesting that we on the front-end filter out all OMIA as non-human, or something else?
The conservative thing is to exclude anything with in_taxon != NCBITaxon:9606 (and allow the user to see these by ticking a "show non-human animal diseases" box
Agreed, that would be fine, but it is my understanding that this taxon distinction is not in the source data to begin with, hence this ticket whose origins reach back to 2015. https://github.com/monarch-initiative/monarch-disease-ontology-RETIRED/issues/59. cc: @TomConlin ?
What is lacking in MONDO is ontological commitment for likely human-only diseases, but anything that is definitively non-human should have an axiom to this effect, e.g.
http://www.ontobee.org/ontology/MONDO?iri=http://purl.obolibrary.org/obo/MONDO_0005748
@cmungall @monicacecilia I do not understand the above discussion. BL API returns the species right? and so why can't we implement my original suggestion:
it might be better to return the results like
human 20
mouse 32
zebrafish 3
etc
and then go to the details --
It would be very useful to show the diseases according to species on the landing page following a search on say "Marfan syndrome."
Originally submitted by @pnrobinson to /monarch-app repo. https://github.com/monarch-initiative/monarch-app/issues/1639