The PHI Phenotype section has the option to filter by High Level Terms, but the Effector high level term doesn't seem to be available for filtering.
For example, for PHIG:344, the following annotation has the 'Effector' high level term:
But 'Effector' can't be selected in the High level term filter:
Note that filtering by Effector is completely pointless in this case, since every PHI Phenotype annotation on the gene page is tagged with 'Effector' (presumably by virtue of the metagenotype containing an Effector gene), so filtering by effectors won't exclude any annotations.
@CuzickA Can you remember why we decided to tag every PHI Phenotype with 'Effector' in this case?
The logic I decided on for tagging PHI Phenotype annotations with Effector in issue #63 was:
the annotation is a metagenotype annotation involving a pathogen gene annotated with GO:0140418 (or any of its child terms).
To me, this implies that the gene should already have an Effector high level term in the GO Biological Process section, meaning the gene page will be picked up by an Advanced Search for 'Effector', and the Effector tags on the PHI Phenotype annotations are redundant from a search perspective.
Even if the gene is not annotated with a GO effector term in one curation session, but is annotated with it in another session, all the annotations will be grouped together on the same gene page anyway, so we still don't need Effector tags on the PHI Phenotype annotations in this case.
The only case I can foresee is if the user tries to do an advanced search on the PHI Phenotype annotation type and the high level term 'Effector', and the search logic uses a combination of these criteria (annotation type and high level term) instead of one or the other (annotation type or high level term). In the former case, the PHI Phenotype annotations won't be picked up unless they have an Effector high level term; in the latter case, they would be.
I might need to confirm with MC how the search logic would behave for combinations of annotation type and high level term, but still, this type of search strikes me as a pretty niche use case.
We decided that we want to keep the Effector high level term in the annotation details, but there's no point adding it to the high level term filter since (for reasons described above) it wouldn't have any effect.
The PHI Phenotype section has the option to filter by High Level Terms, but the Effector high level term doesn't seem to be available for filtering.
For example, for PHIG:344, the following annotation has the 'Effector' high level term:
But 'Effector' can't be selected in the High level term filter:
Note that filtering by Effector is completely pointless in this case, since every PHI Phenotype annotation on the gene page is tagged with 'Effector' (presumably by virtue of the metagenotype containing an Effector gene), so filtering by effectors won't exclude any annotations.
@CuzickA Can you remember why we decided to tag every PHI Phenotype with 'Effector' in this case?
The logic I decided on for tagging PHI Phenotype annotations with Effector in issue #63 was:
To me, this implies that the gene should already have an Effector high level term in the GO Biological Process section, meaning the gene page will be picked up by an Advanced Search for 'Effector', and the Effector tags on the PHI Phenotype annotations are redundant from a search perspective.
Even if the gene is not annotated with a GO effector term in one curation session, but is annotated with it in another session, all the annotations will be grouped together on the same gene page anyway, so we still don't need Effector tags on the PHI Phenotype annotations in this case.
The only case I can foresee is if the user tries to do an advanced search on the PHI Phenotype annotation type and the high level term 'Effector', and the search logic uses a combination of these criteria (annotation type and high level term) instead of one or the other (annotation type or high level term). In the former case, the PHI Phenotype annotations won't be picked up unless they have an Effector high level term; in the latter case, they would be.
I might need to confirm with MC how the search logic would behave for combinations of annotation type and high level term, but still, this type of search strikes me as a pretty niche use case.