TRAPI validation issues with MolePro #878
Labels
external
Requires fixes to an external service
Comments
|
The query graph in question: {
"message": {
"query_graph": {
"nodes": {
"sn": {
"ids": [
"CHEBI:45713"
],
"categories": [
"biolink:SmallMolecule",
"biolink:ChemicalEntity"
]
},
"on": {
"ids": [
"NCBIGene:2739"
],
"categories": [
"biolink:Gene",
"biolink:Protein"
]
},
"un": {
"categories": [
"biolink:NamedThing"
]
}
},
"edges": {
"e2": {
"subject": "sn",
"object": "on",
"predicates": [
"biolink:related_to"
],
"knowledge_type": "inferred"
},
"e0": {
"subject": "sn",
"object": "un",
"predicates": [
"biolink:related_to"
],
"knowledge_type": "inferred"
},
"e1": {
"subject": "un",
"object": "on",
"predicates": [
"biolink:related_to"
],
"knowledge_type": "inferred"
}
}
}
}
}
|
|
I've confirmed that MolePro is returning edges with multiple primary sources: curl -X POST \
"https://molepro-trapi.ci.transltr.io/molepro/trapi/v1.5/query" \
-H "Content-Type: application/json" \
-d '{"message": { "query_graph": { "nodes": { "n0": { "ids": ["CHEBI:45713"] }, "n1": { "ids": ["NCBIGene:5743"] } }, "edges": { "e0": { "subject": "n0", "object": "n1", "predicates": ["biolink:affects"] } } } }}' |
New WarningMolePro is using qualifier values that aren't in the biolink model. See 10/20 automated test run https://arax.ci.transltr.io/?r=2dbf14eb-e50c-47d4-9e25-fb487df1d86a |
|
Are you referring to the Biolink Model documented here? These are the qualifiers listed in Biolink Model that MolePro uses: |
|
The TRAPI validation summary doesn't say. It just shows the not-expected qualifier values, not the Based on my digging in the ARAX-UI/BTE-response I linked:
|
|
Why are we looking at the biolink enums while the qualifiers are listed here? |
|
So, the ARAX Validation Results @colleenXu copied and pasted into this issue on October 21st had omitted monarchinitiative and automat-robokop Edge Qualifier values. The unabridged Results of qualifier values (shown below) suggests that it could be the validator that is problematic, where the validator also reports the other Translator components were providing unresolved 'qualifier_value'. (from the 10/20 automated test run https://arax.ci.transltr.io/?r=2dbf14eb-e50c-47d4-9e25-fb487df1d86a) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Related to #865
Noticed in https://arax.ci.transltr.io/?r=c961069f-36da-4369-a141-3aad9234e5ca.
Downloaded BTE's response bte-ci-pf2-validationProblem.json.zip and used notebook to run TRAPI validation locally
(TRAPI "orange" errors don't prevent ARS from using BTE's response. The "red" critical errors do prevent the ARS.)
Error: invalid KL/AT values
Using
unspecified. The valid value to use isnot_provided.Examples
See edges:
59616f9c131b5816c63ee3fb4d109ee6(KL)3c92b53ccab80dca45b5aa9cef449e18(both)Warning: multiple primary knowledge sources
According to Jackson's post below, this is coming from MolePro, not BTE.
Unsure if this is from BTE's handling (merging KG edges?) or originally from MoleProI can also find this same edge when I query BTE in non-creative-mode for this edge
CHEBI:45713-(affects)->NCBIGene:5743.Edge from KP that triggers this warning
The text was updated successfully, but these errors were encountered: