search

monarch-initiative / monarch-legacy

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

refactor phenoset histogram/distributions owlsim calls #514

Open nlwashington opened 10 years ago

nlwashington commented 10 years ago

i've currently implemented a getHistogram functionality that will take a set of phentoype_associations and generate a histogram object (given a set of phenotype categories with which to classify). this currently calls the summary info method from owlsim. but this should be refactored to call a new function (not yet written) in owlsim that will:

  1. return categorical classifications of a phenotype set (given a set of categories)
  2. list the phenotypes (array?) that are subclasses of each category. (subclasses?) (this would allow dynamic re-querying for drilling down a graph.)

query: a=[P1, P2, P3], r=[C1,C2,C3]

possible blob: results=[{item:{id: C1, label:C1label}, includes : [P1,P2]}, {item: {id : C2, label:C2label}, includes : [P3] }, {item: {id : C3, label : C3label}, includes : []}

we might e able to simply augment the getAnnotationInfo function, because it nicely includes other summary stats (including IC measures), so that if the user wanted to be able to plot the informational properties (instead of just counts) to get a sense of proportion of information a phenotype contributed in the overall set, the IC score could be used to show it.

jmcmurry commented 8 years ago

@nlwashington what is the status with this? Seems like it may have been the sort of thing that got fixed along the way?