We are now using the -base releases of many ontologies as input. This is good because it means we don't load the same terms from multiple sources, but it has the downside that we need to pull in critical dependencies for these ontologies separately. Specifying these here prompts the pipeline to generate a module from the dependent ontology, generated using full ontologies as seeds.
The most critical dependency for our purposes is ro.owl - which provides the semantics of objectProperties (relations) used in the integrated triplestore.
ubyndr
commented
2 years ago
We are now using the -base releases of many ontologies as input. This is good because it means we don't load the same terms from multiple sources, but it has the downside that we need to pull in critical dependencies for these ontologies separately. Specifying these here prompts the pipeline to generate a module from the dependent ontology, generated using full ontologies as seeds. The most critical dependency for our purposes is ro.owl - which provides the semantics of objectProperties (relations) used in the integrated triplestore.
TODO Edit https://github.com/kharchenkolab/cap-pipeline-config/blob/development/config/collectdata/vfb_slices.txt to add http://purl.obolibrary.org/obo/ro.owl