1. MGED Ontology [--usage]
MGED Ontology Working Group
Dec, 2003

2. Outline Brief overview of MGED Ontology (MO)
Structure of MO
What is in MO
Uses of MO
MAGE and MO
MGED Ontology use policies
Simple examples
Nested term examples
Referencing outside terms with MO

3. Ontologies in general Explicit specification of a conceptualization
Defines a vocabulary with which queries and assertions are made
Classes - base type
Animal, Cat, Dog
BioMaterial, Age, TechnologyType
IS-A relationships: Cat IS-A Animal
Properties - attributes of classes
number_of_legs, natural_enemy
HAS-A relationships: Cat HAS 4 legs
Individuals - instantiations of classes
calico, hound, “Stinky Pete”
Axioms - logic statements
Cat != Dog

4. MGED Ontology (MO) Core ontology limited to MAGE-OM v.1.1
Extended Ontology augments the Core with terms currently outside the scope of MAGE v.1.1
OilEd editor:
Protogé: do google search for it

5. MAGE and MO

6. MAGE and MO

7. Main focus of MGED Ontology
Structured and rich description of BioMaterials
BioMaterial
+characteristics
OntologyEntry
+associations

9. A graph representation

10. MGED Ontology Announcements
About to release first stable release of the Core ontology (version 1.1.7)
Provide usage of the ontology and examples
Lot’s of links to ontology resources on the web
Policies
Any term not referenced via DatabaseEntry is by default user-defined and not part of MO
‘accession’ and ‘URI’ are not both needed
Refers to OntologyEntry.value
MO terms may be used in “category” for other ontology’s term (e.g. “Organism”)

11. Referencing MO Terms <MAGE-ML> …
<Database name="The MGED Ontology" identifier="MO"
URI=" />
<OntologyEntry category="SomeMOClass" value= "some_MO_instance">
<OntologyReference_assn>
<DatabaseEntry accession="#some_MO_instance"
URI="
<Database_assnref>
<Database_ref identifier="MO"/>
</Database_assnref>
</DatabaseEntry>
</OntologyReference_assn>
</OntologyEntry>
</MAGE-ML>

12. Graph representation of previous slide’s XML

13. Individuals Instance term goes into value and is the referenced term.
Parent term goes into category

15. Classes with Properties
Nested structure via OntologyEntry.Association
Fillers:
base types (boolean, string, etc.)
Classes/Instances
Enumeration of Instances
OntologyEntry category & value = Property
Filler is given as OntologyEntry Association
follow the rules
When a base type, OntologyEntry.category = Property

16. Measurement Example

17. Enumeration of instances
Base type
Class/Instance

18. Enumerated instances example

19. Base type example

20. Class/Property example

21. This is BAD! An example of node skipping or tree collapsing

22. Including parent classes
You may include parent classes for ease of queries, but you must reproduce all of the structure from the parent to the term of interest

23. Here we have included the parent term “Action” to the instance term “add” and have included the full term tree.

24. “Abstract” classes These are classes without individuals
Usually have “user-defined” instance terms
Term goes in OntologyEntry “category” and “value” attributes and referenced via DatabaseEntry
User-defined term is “Association” (or child OntologyEntry term) with abstract class in category and new term in value, unreferenced
Example: Your BioMaterial source had a Disease “foobar”

26. Referencing terms in other ontologies
Special case of last slide, were new term optionally references some other ontology via its DatabaseEntry
In addition, specify from within the MGED Ontology where the term hailed from
Is-a “OntologyEntry” term has properties:
has_database = source for the term
accesion = accession for the term in
value = the term from the source

27. Without extra DatabaseEntry

28. With DatabaseEntry

29. Thanks
MGED Ontology Working Group