search

VirtualFlyBrain / VFB_json_schema

VFB json schema spec + code for rolling queries producing VFB json schema + related integration tests
Apache License 2.0
0 stars 0 forks source link

Publication query #129

Open dosumis opened 3 months ago

dosumis commented 3 months ago

Component queries:

  1. anatomy to pub NLP + ontology
MATCH p=(pub)<-[r:has_reference]-(n:Neuron {label: "Kenyon cell"}) where pub.title is not null  return pub.title, pub.miniref, n.label, r.typ

  1. curation - anat_2_ep

    MATCH (neuron:Neuron { label: 'Kenyon cell'}) MATCH (ep:Expression_pattern:Class)<-[ar:overlaps|part_of]-(anoni:Individual)-[:INSTANCEOF]->(anat:Class) WHERE anat.short_form in [neuron.short_form]  MATCH (pub:pub { short_form: ar.pub[0]}) RETURN pub.title, pub.miniref

  2. class to individual to dataset to pub 

    MATCH(c:Class:Neuron)-[:INSTANCEOF]-(n:Neuron:Individual)-[:has_source]->(a :DataSet)-[:has_reference]->(r) RETURN DISTINCT c.label, r.label, a.label LIMIT 10

Note - this will add general connectome refs to almost everything.

How to implement in VFB_JSON

  1. We can use existing object rolling in VFB_JSON roll_pub_return("p")
  2. We need a way to to aggregate and unique results from With 3 different queries, we need a way to aggregate and unique
  3. Ideally we should extend VFB_JSON to support origin (so we can flag NLP
  4. Consider using ontology closure.