Annotation Conf. Call May 28, 2013
Jump to navigation Jump to search
- New GO_REF for apoptosis IC
- Auto repair of annotations to obsolete and merged terms- Plan is to have Mike's filtering script transfer annotations when there is a one-to-one replacement of obsolete and merged terms. Any questions?
- URL for PANTHER IDs- http://pantree.org/node/annotationNode.jsp?id=PTN000177279
- If you are loading/displaying PAINT annotations, please have the PANTHER IDs in the With column point to this URL. This leads one to the evidence trail.
- Discussion of Database Identifiers Used in the With/From Field
- Identifiers are required in the With/From field for some evidence codes, such as IPI and IGI.
- For other evidence codes, such as IMP, entries in the With/From field are optional.
- For developing the Common Annotation Framework, we would like to review what identifiers are used in the With/From field and, for some evidence codes like IMP, what the intended use of the information would be.
- All entities included in the With/From field need to use a valid database identifier. What is the best way to track the syntax and validity of identifiers curators want to place in the With/From field?
- WB has database identifiers for Variations (alleles), RNAi experiments, Regulatory interactions, Expression patterns, etc.
- In the past, WB curators have used Variation, RNAi and Phenotype Ontology term IDs in the With/From field for IMP annotations to provide users a more direct link between a phenotypic observation and the GO term selected. Is this useful to consumers of the annotation file or is it just useful internally for tracking the annotation process? Would more direct associations between GO Biological Process terms and Phenotype Ontology terms help in evaluating annotation consistency and provide training data for semi-automated BP annotation?
- In some cases, the information that WB is adding to Column 16 annotation extensions is already contained within an existing database object. Since the Column 16 information has thus already been curated in another object, we'd like to explore the possibility of semi-automated population of Column 16 data. How important is it to record the database object from which the Column 16 information was derived? Again, is this only useful internally for curation purposes or is it also helpful to users?