GO annotation fields required for correct display: Difference between revisions
m (→COMMENTS) |
mNo edit summary |
||
Line 3: | Line 3: | ||
GO annotations are most completely represented by the gene association file format used by GO Consortium members, detailed [http://www.geneontology.org/GO.format.gaf-2_0.shtml#fields here]. | GO annotations are most completely represented by the gene association file format used by GO Consortium members, detailed [http://www.geneontology.org/GO.format.gaf-2_0.shtml#fields here]. | ||
However, for a web-based display of annotation data it is understandable | However, for a web-based display of annotation data it is understandable that groups may want to restrict their display to the most essential and useful GO annotation fields required by their users. However, the GO Consortium expects that displays of GO annotation data to include the mininimal fields which are essential to be able to properly interpret a GO annotation statement, and ideally to provide recognition of the date annotation created and the data source. | ||
In addition, we would strongly encourage web sites to provide a link to the GO Consortium, so that users can further explore GO resources. | In addition, we would strongly encourage web sites to provide a link to the GO Consortium, so that users can further explore GO resources. |
Revision as of 04:59, 10 August 2011
Recommendations for the Correct Display of Gene Ontology Data
GO annotations are most completely represented by the gene association file format used by GO Consortium members, detailed here.
However, for a web-based display of annotation data it is understandable that groups may want to restrict their display to the most essential and useful GO annotation fields required by their users. However, the GO Consortium expects that displays of GO annotation data to include the mininimal fields which are essential to be able to properly interpret a GO annotation statement, and ideally to provide recognition of the date annotation created and the data source.
In addition, we would strongly encourage web sites to provide a link to the GO Consortium, so that users can further explore GO resources.
If groups would like to be notified of any changes to these recommendations, then please get in touch and subscribe to the GOfriends e-mail list.
Recommendations for provision of searching facilities for GO terms
Web sites providing users with a way of filtering and viewing GO annotations or terms should take into account the the structure of the Gene Ontology, which identifies descendant terms. For instance, where a user is looking for annotations to 'nucleus', should bring back terms or annotations not only corresponding to the directly searched term 'GO:0005634; nucleus', but also to all descendant terms (for example:'GO:0031039 : macronucleus', 'GO:0016363 : nuclear matrix').
One resource that can assist groups to interrogate the ontologies and reason over relationships is OWLtools.
Column No. | Column Name | Requirement | Further Detail |
---|---|---|---|
1. | DB | Optional | |
2. | DB_Object_ID | Essential | can be mapped to corresponding sequence identifiers |
3. | DB_Object_Symbol | Optional | |
4. | Qualifier | Essential | Notably annotations using the 'NOT' qualifier |
5. | GO term ID and/or name | Essential | |
6. | DB:Reference | Essential? Recommended? | this is an integral part of the annotation format. Without the reference, the annotation becomes untraceable and causes problems for users of the displayed annotation set who would like to verify the statement |
7. | Evidence | Essential | |
8. | With(or)From | Recommended | Provides important supporting information, especially required to add meaning to annotations describing interactions |
9. | Aspect | Optional | |
10. | DB_Object_Name | Optional | |
11. | DB_Object_Synonym | Optional | |
12. | DB_Object_Type | Optional | |
13. | Taxon | Optional | |
14. | Date | Essential | |
15. | Assigned_by | Recommended/Essential? | |
16. | Annotation Extension | Optional | |
17. | Gene Product Form ID | Optional |
COMMENTS
1. In addition, should external websites be encouraged to provide a link to the GO Consortium? Rachael
Have added into draft Edimmer
2. we could also suggest that if external websites would like to be notified of any changes to these recommendations, then to send us an email? We would then be able to build up a picture of who displays our annotations, which may be useful for grant reporting purposes. Edimmer 11:12, 2 February 2011 (UTC)
- Have added into draft Edimmer
3. Suggested change to the second part of the intro to this statement: The display of annotation data by some web-based databases is often restrict to the most essential and useful GO annotation fields required by their users. However, the GOC expects the display of all GO data to include the mininimal fields which are essential to enable appropriate interpretion of a GO annotation statement, and also to provide recognition of the data source. Ruth]
- Have added into draft Edimmer
4. Should also state that websites providing users with a way of filtering and viewing GO annotations or terms should take into account the GO structure. i.e. therefore when a user is looking for annotations to 'nucleus', should bring back annotations not only to the directly searched term, but also to descendants (nucleolus, nuclear membrane). GO managers call feedback.
5. Tanya: I would prefer that the date is essential and the assigned by is recommended (rather than essential). I think it's more important to know how old an annotation is than to know who made it.
- Have added into draft, although perhaps whether to settle on essential/optional for the assigned_by field needs further discussion? Edimmer
Examples of external website displays
1. Ensembl e.g. Stat6
Displays GO ID, GO term name, Evidence code, 'with' field information (only seems to be for Compara annotations, incomplete for other IEA methods, ISS and IPI).
2. NCBI
e.g. Stat6
Displays GO term name, evidence code, PMID link