Support for slims in OLS

[simonjupp]

Slims and categories in OLS for exposing application ontologies e.g. atlas, scAtlas, HCA, GWAS. OLS supports slims for search, but not for filtering against the REST API. It is also not possible to view slims. Slims are useful for downstream applications and consumers. In particular we want to use slims as a mechanism for delivering API access to subsets of terms in EFO e.g. Array have a specific checklist of terms

[henrietteharmse]

In discussion on Slack @matentzn mentioned this example of slims showing that an ontology can have up to 50 slims:

format-version: 1.2
data-version: fbbt/releases/2020-10-15/fbbt-non-classified.owl
date: 15:10:2020 14:37
subsetdef: abnormal_slim "Abnormal/normal slim"
subsetdef: attribute_slim "Attribute slim"
subsetdef: BrainName "BrainName project terms"
subsetdef: BrainName_Embargo "Terms to hold back pending BrainName paper publication"
subsetdef: cell_quality "cell_quality"
subsetdef: Cross-product_review "Involved_in"
subsetdef: cur "cur"
subsetdef: deprecated "warning of impending obsoletion"
subsetdef: EmbDevSlim "EmbDevSlim"
subsetdef: expressionslim_flybase_ribbon "FlyBase ribbon terms"
subsetdef: FB_gloss "terms for FlyBase glossary page"
subsetdef: FlyTed "Testis slim"
subsetdef: FT_new "FlyTed temp subset for edit tracking purposes"
subsetdef: goantislim_grouping "Grouping classes that can be excluded"
subsetdef: gocheck_do_not_annotate "Term not to be used for direct annotation"
subsetdef: gocheck_do_not_manually_annotate "Term not to be used for direct manual annotation"
subsetdef: goslim_agr "AGR slim"
subsetdef: goslim_aspergillus "Aspergillus GO slim"
subsetdef: goslim_candida "Candida GO slim"
subsetdef: goslim_ChEMBL "ChEMBL protein targets summary"
subsetdef: goslim_chembl "ChEMBL protein targets summary"
subsetdef: goslim_drosophila "Drosophila GO slim"
subsetdef: goslim_flybase_ribbon "FlyBase Drosophila GO ribbon slim"
subsetdef: goslim_generic "Generic GO slim"
subsetdef: goslim_goa "GOA and proteome slim"
subsetdef: goslim_metagenomics "Metagenomics GO slim"
subsetdef: goslim_mouse "Mouse GO slim"
subsetdef: goslim_pir "PIR GO slim"
subsetdef: goslim_plant "Plant GO slim"
subsetdef: goslim_pombe "Fission yeast GO slim"
subsetdef: goslim_synapse "synapse GO slim"
subsetdef: goslim_virus "Viral GO slim"
subsetdef: goslim_yeast "Yeast GO slim"
subsetdef: gosubset_prok "Prokaryotic GO subset"
subsetdef: hidden_assertion "Stanza contains hidden assertion."
subsetdef: high_level_annotation_qc "High-level terms not to be used for direct annotation"
subsetdef: hpo_slim "Human phenotype slim"
subsetdef: larval_OF "larval olfactory system"
subsetdef: mf_needs_review "Catalytic activity terms in need of attention"
subsetdef: mpath_slim "Pathology slim"
subsetdef: olfactory_system_draft "adult olfactory system model"
subsetdef: RD "terms for rd test ontology"
subsetdef: ro-eco ""
subsetdef: RO:0002259 ""
subsetdef: scalar_slim "Scalar slim"
subsetdef: termgenie_unvetted "Terms created by TermGenie that do not follow a template and require additional vetting by editors"
subsetdef: TL_lineage_clone "Lineage clone related term from Tzumin Lee dataset"
subsetdef: value_slim "Value slim"
subsetdef: virus_checked "Viral overhaul terms"

[henrietteharmse]

In conversation with @zoependlington she mentioned an EFO slim they have for Open Targets. Interestingly in this use case the Open Targets slim of EFO is not a subset of EFO. I.e., it includes terms that are not in EFO, but that have been added to the Open Targets slim. The terms that are in the Open Targets slim of EFO that are not in EFO are added because these are terms that are under discussion to potentially be added to EFO.