Include other methods to download ontology content and convert to a SQLite database using semsql

[tursynay]

[tursynay]

Rather than relying on the set of existing ontologies provided as semsql databases that can then be used with OAK, use the BioPortal Term search API (https://data.bioontology.org/documentation#nav_search) in cases where a semsql database does not already exist for the ontology content of interest.

• Trish - check if there are any limits on number (thousands of records) of API calls that can be made to the Term search API. If there is a limit, Trish will include into the documentation and add options to "chunk" the search inputs or other ways to work with the API.
  • Email sent

[tursynay]

@twhetzel, following up on this one.

[twhetzel]

BioPortal API limits: The only limitation we impose on API requests is throttling based on request type. For lightweight calls (term search likely falls into this category), the limit is 15 requests per second per IP address. For more resource-intensive endpoints, such as the /annotator and /recommender services, the limit is 3 requests per second per IP.

The BioPortal API should work so I would not add extra complexity to have a local database version for any content not already represented as a SQLite database available to OAK and access these other ontologies via the BioPortal API.