Gene Ontology in Pathway Tools: Internals

1
Gene Ontology in Pathway Tools Internals
2
Relationship between GO and a PGDB

• GO is stored in its own Ocelot KB
• Class-subclass relationship defines the GO is-a
  hierarchy
• Additional slots PARTS, PART-OF define the GO
  part-of hierarchy
• Each PGDB has a class Gene-Ontology-Terms
• Slots GO-TERMS, TERM-MEMBERS encode annotations
• When annotated a gene product to a GO term, term
  and its parents are imported into the PGDB
• Searching for gene/product by GO term only
  requires access to terms in PGDB
• Editing GO terms for gene/product requires access
  to full GO KB

3
Where does GO Ocelot KB come from?

• Parsing tool generates Ocelot KB from .obo-xml
• GO KB is built into Pathway Tools runtime
  distributions
• BUT
• GO is updated much more frequently than Pathway
  Tools software releases
• How to incorporate new terms into a PGDB?

4
Updating GO Ocelot KB

• Download new .obo-xml file
• (update-go-kb input-file ltfilenamegt)
• Checks to see if input-file is more recent than
  currently installed GO KB
• Saves file aic-export/go/beta/go.ocelot
• At startup time, Pathway Tools looks for file and
  loads it, overriding GO KB in memory
• Consistency Checker validates PGDB to match the
  GO KB
• Updates hierarchy, importing new terms as
  necessary
• Deletes terms w/ no subs, parts, or annotated
  gene products
• Reports on obsolete terms that have gene products
  annotated to them curator should reassign these