Closed frostyfan109 closed 5 years ago
This does not seem like this is what's happening. The query planner generated this plan using the query Workflow 5 v3:
[SELECT {'population_of_individual_organisms': population_of_individual_organisms:[], 'chemical_substance': chemical_substance:[]} | [edge[dir:->,pred:None]] from:/clinical/cohort/disease_to_chemical_exposure where:[['table', '=', 'patient'], ['table', '=', 'patient'], ['year', '=', 2010], ['year', '=', 2010], ['cohort_features.AgeStudyStart', '=', '0-2'], ['cohort_features.AgeStudyStart', '=', '0-2'], ['feature.EstResidentialDensity', '<', 1], ['feature.EstResidentialDensity', '<', 1], ['maximum_p_value', '=', 1], ['maximum_p_value', '=', 1]] set:[], SELECT {'chemical_substance': chemical_substance:[], 'gene': gene:[]} | [edge[dir:->,pred:None]] from:/graph/indigo where:[] set:[], SELECT {'chemical_substance': chemical_substance:[], 'gene': gene:[]} | [edge[dir:->,pred:None]] from:/graph/gamma/quick where:[] set:[]]
Both Indigo and Robokop are being queried about chemical_substance->gene.
Now the issue is, if Indigo returns no results, it will halt the query. This is because Indigo does not return a knowledge map, so the query cannot continue.
All statements that have the same query need to share the nodes from the collective previous statement. They should not be using the previous statement's nodes as is it not the actual previous statement.
Ex: Schema1: chemical_substance->treats->disease Schema2: chemical_substance->treats->disease
Schema2 would be the only source that is queried.
Not sure if this actually is happening, but it seems like it is. Needs to be fixed if it is indeed occurring.