search

monarch-initiative / monarch-legacy

Monarch web application and API
BSD 3-Clause "New" or "Revised" License
42 stars 37 forks source link

Create prototype patient page #770

Open cmungall opened 9 years ago

cmungall commented 9 years ago
  1. load test data or subset of test data from @mellybelly into SciGraph, cc @nlwashington
  2. create a new category, e.g. 'patient' (or could be more general), @ccondit can change the yaml conf
  3. ensure this passes through to golr. Should be trivial this is just analogous to a disease-pheno association
  4. make a page, /patient/:id, by default may look exactly like a disease page
cmungall commented 9 years ago

We will do just phenotypes and gene as a first pass.

@nlwashington the email has subject 'UDP patient data for Monarch' June 2

nlwashington commented 9 years ago

i also have this kind of data from coriell... but i was not materializing the "person" identifiers. should i?

harryhoch commented 9 years ago

do we have to deal with any demographic data?

kshefchek commented 9 years ago

Is there anything I can do currently or are we waiting on data ingest and/or GOlr loading updates?

kshefchek commented 9 years ago

cc @nlwashington and @cmungall, do we still want this for the upcoming release?

cmungall commented 9 years ago

the timing can be decoupled from the release (we can demo it on beta), but still high priority.

nlwashington commented 9 years ago

to do this, we need to:

  1. create new categories (type by foaf:person).
  2. create new association documents in golr for patient-to-:
    • phenotype
    • genotype
    • gene
    • variant
    • models
    • disease
    • (i am not sure how the other categories would map yet: drugs, pathways)
kshefchek commented 9 years ago

Very prototypey pages added in /labs/case, for example: http://beta.monarchinitiative.org/labs/case/_:genid-nodeid-_:person-NINDS2400-1 http://beta.monarchinitiative.org/labs/case/_:genid-nodeid-_:person-NINDS3391-1 http://beta.monarchinitiative.org/labs/case/_:genid-nodeid-_:person-691-40

nlwashington commented 9 years ago

this one has a genotype (but is not labeled or linked): http://beta.monarchinitiative.org/labs/case/_:genid1671454 should have this label: G6PD(SER188PHE); 47;Y;t(X;7)(Xpter>Xq24::7q32>7qter; 7pter>7q32::Xq24>Xqter);+X [GM00324] could link here: http://beta.monarchinitiative.org/genotype/dbSNPIndividual:17729

similarly: http://beta.monarchinitiative.org/labs/case/_:genid-nodeid-_:person-222-1 should have something like: GLB1;GLB1(ARG482HIS;TRP273LEU) [GM01602]

harryhoch commented 9 years ago

are we going to want to pull-over any of the details found on the coriell pages?

jmcmurry commented 8 years ago

Case is the right level of abstraction; one or more patients can be in a case. Case will be a special form of disease. We also need some place to put "demographic" data (what is the right term for that in animals?).

We need to think about the model by which we represent the continuum of taxa, families, individual patients, patients of the same disease cohort, diseases, derived biosamples from individuals, cell lines etc. @mbrush please advise. @mellybelly said you did some of this modeling for eagle-i? Could we resurrect it?

kshefchek commented 8 years ago

Dipper script and cypher queries are ready to go, reassigning to @DoctorBud. Feel free to assign back if you don't have the bandwidth.