Page History
...
Here we present concepts that are specific to our application, and terms used to refer to them. Concepts and terms established in semantic web / ontology communities and standards are taken to be known by users, and not defined here.
Concept 1: Domain vs Application Ontology Content
"Domain" content in the eagle-i ontologies includes all axioms representing general knowledge pertinent to the domain of research resources. This content is partitioned into a set of files referred to as the "domain layer" of the ontology, which comprise the publicly released ERO ontology module that is shared for community re-use.
...
| Info | ||
|---|---|---|
| ||
See the Table of Annotation and Domain Layer Files for a list of IRIs, key features and content of files that comprise the domain and application layers. |
Concept 2: Extended vs Core Domain Content (model / files / classes)
This is an artificial distinction made based on how MIREOTed content is stored and loaded into memory.
"Extended" content (files / taxonomies / terms) comes from five ontology imports (Uberon, GO, MP, Pato, DOID). Due to their size, these are stored in owl files separate from the ero.owl core file that holds all other domain content. Their content makes these files too large to be loaded into memory as the application runs, and are therefore indexed separately and dynamically called as needed. These files are also handled separately by build scripts and housed in separate files at the end of the build pipeline.
...
The extended vs core distinction is important, as specific application annotations are needed to guide the software in finding axioms in the right files, as detailed in the Table of Annotation Properties and in the Task Workflow and Scenarios below. The classification of each file in the eagle-i ontology as being extended vs core is indicated in the far right column of the Table of Annotation and Domain Layer Files .
Concept 3: Resource Classes (aka Instantiated Classes)
“Resource” classes represent entities considered to be research resources for which instances are created in the eagle-i data. Subtypes of this concept include "primary" resources that are cataloged and shared, "embedded" resources that are directly related (1:1) with a single primary resource instance, and "stubbed" resources that can be related to more than one primary or embedded resource instance (and also not cataloged or shared in eagle-i).
| Class Type | Description | Examples |
|---|---|---|
| Primary Resource Classes | Represents the main resources types cataloged and shared as instances in the eagle-i data. | Examples include: 'antibody reagent', 'software', 'service offering', 'instrument', ‘cell line’, etc. |
| Embedded Resource Classes | Represents non-primary resources for which instances are created and that are linked 1:1 with a single resource instance. Embedded resources are not cataloged and shared by eagle-i. | Examples include: 'construct inserts' as embedded resource types for 'DNA constructs', 'phenotype annotations' as embedded resource types for 'organisms'. |
| Stubbed Resource Classes | Represents non-primary resources for which instances are created that can be linked to more than one resource instance. Stubbed resources are also not cataloged and shared by eagle-i. | Examples include: 'genetic alteration’ instances, which can be associated with both primary resources or other stubbed resources (such as a ‘cell line’ instance and a ‘human subject’), and 'human subject' instances, which can be linked to more than one 'primary cell line' instance. |
Concept 4: Referenced Taxonomy
Referenced Taxonomies are ontology hierarchies that hold non-resource classes used as values to describe resource attributes. In the data, the IRIs of referenced taxonomy classes are used directly as the objects of RDF triples about a resource instance. Instances of referenced taxonomy classes are not created in the eagle-i data.
...
Finally, it is also important to have a basic understanding of the file import structure through with the complete eagle-i ontology is assembled in order to correctly place axioms driving application functionality. The Table of Annotation and Domain Layer Files above describes key features and content of each file. Note that each owl file containing domain content is paired with an application file (-app.owl) that holds application axioms on domain classes/properties. See the Ontology File Structure page for more details and a diagram of the import chain.
...
Overview
Content Tools