Annotation Conf. Call 2016-03-22
Bluejeans
Agenda
IC from ISS
- From 2016-03-08 annotation call, we agreed that IC annotations from existing ISS annotations would be okay, but would require a new GO_REF to document the annotation practice.
- Rebecca drafted the text of the new GO_REF:
GO_REF:0000109 title: Gene Ontology annotations Inferred by Curator (IC) using at least one Inferred by Sequence Similarity (ISS) annotation to support the inference year: 2016 The Gene Ontology Consortium uses the IC (Inferred by Curator) evidence code when assignment of a GO term cannot be supported by direct experimental or sequence-based evidence, but can, based on a curator’s biological knowledge, be reasonably inferred from existing GO annotations to the same gene/gene product. Use of the IC evidence code with GO_REF:0000109 indicates that a curator inferred the GO term based on at least one supporting annotation with an ‘Inferred from Sequence Similarity’ (ISS) evidence code. Note that additional supporting annotations may be experimentally evidenced. When using GO_REF:0000109, the 'with/from' field must contain all GO identifiers used as supporting annotations.
- As well as a proposed update to the GO website documentation:
In cases where the GO term is inferred from sequence similarity evidence, such IC-annotations will use reference GO_REF:0000105. When using this reference, one or more supporting annotations will be ISS-evidenced; a supporting ISS-evidenced annotation may also be used together with a experimentally-evidenced supporting annotation (e.g. IDA or IMP). GO_REF:0000105 can be used where the supporting ISS annotations are transferred from different species. For example, an IC-evidenced GO term can be created for a human protein based on two different GO terms; one ISS'd from mouse to human, and one ISS'd from rat to human.
Use of has_participant Annotation Extension Relation
- Definition: Identifies an entity affected by the gene product's participation in a molecular function or biological process
- has_participant documention on GO wiki
- Okay to use this annotation extension relation when it is capturing legacy information?
- In MGI we captured thousands of annotations where a term was associated with an anatomical structure before the method of using relations was in place. In these cases, we know that the structure somehow participates in the process, but we don't know exactly how. We would like to allow these annotations extensions to pass the checks.
http://amigo.geneontology.org/amigo/gene_product/MGI:MGI:1276112
- Would it be possible to allow our legacy annotations, but have the rule that this relation should not be used for new manual annotations?
Annotation Consistency Exercise
Gene/Marker Name | GO term | Evidence Code | With/From | Annotation Extension | Annotation Extension | Annotation Extension | Annotation Extension | Comment |
---|---|---|---|---|---|---|---|---|
SYD-2 | kinesin binding (GO:0019894) | IPI | UNC-104 | |||||
UNC-104 | protein binding (GO:0005515) | IPI | SYD-2 | |||||
SYD-2 | kinesin binding (GO:0019894) | IDA | has_input (UNC-104) | |||||
SYD-2 | kinesin binding (GO:0019894) | IDA | has_direct_input (UNC-104) | |||||
SYD-2 | kinesin binding (GO:0019894) | IDA | has_input (UNC-104) | part_of (protein binding, bridging GO:0030674) | ||||
UNC-104 | ATP-dependent microtubule motor activity, plus-end-directed (GO:0008574) | IDA |