updating architecture guidelines with qualifiers

[sierra-moxon]

No description provided.

[CaseyTa]

Is there any difference in the qualifiers mentioned in item 2 vs item 4?

[edeutsch]

1. Making identifiers more specific, e.g. responding to a query involving an entity with information related to a subclass of that entity. In practice, the more-specific ID is a node linked to the less-specific node by a subclass_of edge in the knowledge graph. In the knowledge_graph portion of the response, the more-specific identifier must be present and linked to the less-specific identifier. In the results portion of the response, the more-specific response node will be bound to the less-specific query node.

Perhaps this is too pedantic, but I'm thinking that identifiers can't be more specific; they are identifiers. only entities (or concepts or nodes?) can be made more specific? Perhaps this refinement is more accurate:

1. Making entities more specific, e.g. responding to a query involving an entity with information related to a subclass of that entity. In practice, the more-specific entity is a node linked to the less-specific node by a subclass_of edge in the knowledge graph. In the knowledge_graph portion of the response, the more-specific node must be present and linked to the less-specific node. In the results portion of the response, the more-specific response node will be bound to the less-specific query node.

This also seems to imply that if my query involved concept_X and I only have an edge for child_concept_of_X, I MUST still include concept_X in my KG even if I don't really have much to say about it. I MUST NOT just have child_concept_of_X bound to the query node? Is that really what we want to require? If so, maybe this could be a bit more explicit since I am not confident in this inference.

[suihuang-ISB]

Also, perhaps on the pedantic side... Agree with Eric that ID are just that, ids, that is: atomic units.
But there is also a problem with the term "qualifier" in terms of making them "more specific".

2. Making categories and qualifiers in a query more specific. e.g. responding to a query for a biolink:NamedThing with a particular biolink:ChemicalEntity.

We at SPOKE had originally used the term 'qualifier' for indicating the CONDITIONAL validity, Eg: "Compound X activates Gene G" BUT: with the qualifier that this is is only true in TISSUE T. Or in the case of correlations/cooccurrences in clinical/multiomics data, A and B are correlated, but only in the cohort of the elderly (age > X)... etc.
Over time, I think 'qualifier' has evolved away from this original meaning to indicate a more granular specification. down the tree of sub-types (as evident in the recent discussions. )

These two meanings of 'qualifier', that is, CONDITION for validity of a predicate vs. finer-grained SPEIFICATION of said predicate, are fundamentally different from the user perspective, and affect the query. May be useful to say what we mean by 'qualifier' here - and perhaps structurally encode this distinction??