NCATSTranslator / TranslatorTechnicalDocumentation

Official Developer Documentation repository for the Biomedical Data Translator
https://ncatstranslator.github.io/TranslatorTechnicalDocumentation/
Creative Commons Zero v1.0 Universal
3 stars 17 forks source link

add information on what resources each KP/ARA consumes #96

Open andrewsu opened 2 weeks ago

andrewsu commented 2 weeks ago

Currently within Translator, we don't have a great way to track which KPs/ARAs consume which resources (for example, who consumes infores:huri?). We have proposed adding this to the infores catalog repo. The goal is to generate a figure that we will add to the documentation here. (The link goes to a sankey diagram, but probably we'll need to adjust that to something that will scale better.)

We prototyped this process with Eric D and Jackson C on ARAX, RTX-KG2, BTE, and Service Provider, so now asking for info from other ARAs and KPs. Please assign this issue to relevant people.

There are two options on how to submit this info:

Method 1: Create a PR with a script (preferred)

In https://github.com/biolink/information-resource-registry/tree/main/src/information_resource_registry/relation-map/data, you will find several JSON files that look like this (from arax_infores_list.json):

{
  "infores:arax": [
    "infores:answer-coalesce",
    "infores:automat-binding-db",
    "infores:automat-cam-kp",
    "infores:automat-ctd",
    "infores:automat-drug-central",
    "infores:automat-genome-alliance",
...

The interpretation of this file is that infores:arax consumes data from infores:answer-coalesce, infores:automat-binding-db, and the rest of the resources listed. This file is generated by https://github.com/biolink/information-resource-registry/blob/main/src/information_resource_registry/relation-map/src/arax-kps.py, which programmatically queries the ARAX API for this info. This is the preferred method because this script can be set up as a github action, so we can keep this info up-to-date.

Method 2: Submit a PR directly on infores_catalog.yaml

Ultimately the info in the JSON files will be integrated into infores_catalog.yaml, so if you want to keep your info up-to-date manually, you can create a PR on that file.

eKathleenCarter commented 1 day ago

@cbizon @EvanDietzMorris see my PR biolink/information-resource-registry#33 maybe @YaphetKG could suggest how to get the metadata for each source from smart-api rather than automat