GO anotation at a certain level of specificity

Question

[martinaweil]

Hi! im a undergrad helping with a project. Ive annotated a bunch of transcripts using trinotate and have then used go seq for go term enrichement. My thesis director has told me he doesnt want overly specific terms to appear since hed like to make a venn diagram with many species, and wants to show broad similarities. But the output i have had redundant goterms. Is there anyway i can choose the level of go terms? Ive done it in DAVID, but sadly i cant do it for this species (they arent as well researched). Either on trinotate or goseq, is there a way? thank you!

Answer 1

[Jean-Karim Heriche]

You can't simply cut the ontology at a given level if you're trying to get terms of similar semantic specificity. This is because the semantic specificity varies between branches. For example, the term cytokinesis is level 4 down the path cellular process > cell division but level 5 if you go cellular process > cell cycle. There are two ways of dealing with this. The first and easiest is to simply pick relevant terms as a flat controlled vocabulary and assign genes to these terms if they are annotated with any child term. The other is to build a slim ontology, i.e. a simplified subset of the full ontology that focuses on the relevant parts.