Replace NodeNorm/NodeAnnotator with a single annotator

[gaurav]

There are three ways to do this:

1. Bring NodeAnnotator in sync with NodeNorm in terms of identifiers (i.e. normalize queries before responding?)
   • Both systems effectively construct JSON, so once we've chosen the best database store, we could merge all the JSON outputs from the two systems seamlessly if they are using the same canonical identifiers.
2. Provide some way to export NodeAnnotator so that it can be directly incorporated into NodeNorm.
   • This might require precalculating annotations for every node in NodeNorm.
   • This will ALSO require assembling all these properties on-the-fly, which might slow things down (even if we assume that not every query with have the "annotations" flag turned on).
     • Although we could precompute these assembled annotations also.
3. Incorporate NodeAnnotator's pipelines into Babel so that we can calculate the annotations ourselves.
   • "Merge the code" option.
   • Step 1 would help with this anyway.

Note that NodeNorm has its own annotations that need to be consolidated (NCATSTranslator/NodeNormalization#306)

[gaurav]

As per our meeting on 2025mar10, we decided that option 1 was a good place to start, and that option 3 was superior to option 1 in terms of simplicity.

To achieve option 1, the Node Annotator folks are working on the following:

1. Download NodeNorm files and load them into ElasticSearch (this can also be the ElasticSearch NodeNorm for our comparison tests!)
2. Generate a report about what’s consistent and what’s not consistent.
3. In general, Annotator should follow NodeNorm IDs, but if there are cases where we find Annotator has better IDs, we should incorporate that into NodeNorm.

@newgene Could you please assign this issue to Johnathan and Dylan? Thank you!

[newgene]

Looks like I don't have the permission to assign issues. But I can just tag @ctrl-schaff and @DylanWelzel here.