1 of 4

Updating SSSOM-py to use the Bioregistry and curies

2023 Workshop on Prefixes, CURIEs, and IRIs

November 27th, 2023��Nico Matentzoglu, Harshad Hegde, Charles Tapley Hoyt

1

2 of 4

SSSOM Lightning Review

  • Encodes semantic mappings with detailed and precise metadata
  • Entities are referenced by compact URIs (CURIEs)
  • CURIEs require prefix maps for:
    • Validation (i.e., is the prefix defined?)
    • Expansion into URIs for usage in semantic web applications
  • Pain points:
    • Knowing what to put in my prefix maps is a lot of work and error-prone
    • I need to standardize a variety of different prefixes representing the same thing
    • I need to be able to reconcile conflicts and collisions
    • I need to be able to expand and compress CURIEs and URIs
    • I don't want to have to keep implementing this over and over

2

3 of 4

Highlights

  • Defined a prescriptive context in the Bioregistry (https://bioregistry.io/context/obo)
    • Prioritizes OBO Foundry preferred prefixes over Bioregistry canonical prefixes
    • Uses CURIE prefix reconciliation to manually override prefixes for special cases
    • Uses rewiring to manually override URI prefixes for special cases
  • Uses this context as the (reasonable) default when none given
  • Reimplemented prefix map I/O, chaining, reconciliation, expansion, compression, and standardization operations using the curies Python package
  • Simplified downstream usage of SSSOM in the Ontology Access Kit (OAK)

3

4 of 4

Pull Requests

Many pull requests between September and October 2023 made this happen:

4