Does the GAF gene ontologies file only contain terms that are leaf nodes?
When I query the file to get all GO biological process terms associated with a particular gene, I was surprised when I did not get back the "Biological Process" root term as technically all genes in this particular ontology should belong in this term (due to the hierarchical structure of GO, all terms propagate their genes to their parent nodes, ulitmately resulting in the root node having all gene annotations). Then I looked online and found this question where in the comments it says that it is in fact the case that in particular for the GAF file (e.g. goa_human.gaf), only the leaf nodes are considered. Is this documented somewhere? I could not find anything in GO pages. It seems like quite an important detail to be left out of the description of the file format. Can anyone confirm this and preferably add a reference?
• 22 views