I'd like to alert you to upcoming changes in the way taxon constraints are included within the 'go-plus' version of the ontology. A taxon constraint is a relation from a GO term to an NCBI taxon identifier. There are two types of taxon constraints currently included in the ontology: 'only in' and 'never in'. The taxon constraints asserted by GO editors have been included within go-plus.owl for the last few years. However a larger number of effective constraints are used within the ontology and annotation QC; these are logically implied by relations from GO terms to other ontology terms for anatomy and cell types, which come with their own taxon constraints in imported ontologies such as Uberon and CL.
In the upcoming December release, we will include precomputed inferred taxon constraints for each GO term, whether asserted by GO editors or inferred from relations to external terms or to other GO terms. This will provide a more complete resource for the effective taxon constraints for each GO term. Computed taxon constraints will use the relations 'in taxon' (RO:0002162) and 'never in taxon' (RO:0002161). Existing assertions using 'only in taxon' (RO:0002160) will be unchanged.
In the January release, the relation used for 'only in' taxon constraints will change from RO:0002160 'only in taxon' to its parent RO:0002162 'in taxon'. RO:0002160 is planned for obsoletion within RO. In practice these two relations have the same meaning (all reasoning on taxon constraints is already performed in terms of 'in taxon').
Hi all,
I'd like to alert you to upcoming changes in the way taxon constraints are included within the 'go-plus' version of the ontology. A taxon constraint is a relation from a GO term to an NCBI taxon identifier. There are two types of taxon constraints currently included in the ontology: 'only in' and 'never in'. The taxon constraints asserted by GO editors have been included within go-plus.owl for the last few years. However a larger number of effective constraints are used within the ontology and annotation QC; these are logically implied by relations from GO terms to other ontology terms for anatomy and cell types, which come with their own taxon constraints in imported ontologies such as Uberon and CL.
In the upcoming December release, we will include precomputed inferred taxon constraints for each GO term, whether asserted by GO editors or inferred from relations to external terms or to other GO terms. This will provide a more complete resource for the effective taxon constraints for each GO term. Computed taxon constraints will use the relations 'in taxon' (RO:0002162) and 'never in taxon' (RO:0002161). Existing assertions using 'only in taxon' (RO:0002160) will be unchanged.
In the January release, the relation used for 'only in' taxon constraints will change from RO:0002160 'only in taxon' to its parent RO:0002162 'in taxon'. RO:0002160 is planned for obsoletion within RO. In practice these two relations have the same meaning (all reasoning on taxon constraints is already performed in terms of 'in taxon').
For more information on how taxon constraints are used for QC in OBO ontologies, see https://oboacademy.github.io/obook/explanation/taxon-constraints-explainer/
These changes can be previewed in the current SNAPSHOT version of go-plus.owl.