Is there any way to group GO term (biological process) based on close functionality

Question

[Space_Life]

Hi, I have hundreds of GO terms (biological process) which I want to bundle together to make it easier to explain. Is it okay to do it? How do you categorize GO terms? Is there any already established method? I would appreciate any help. Thank you.

[cpad0112]

you can go a level up in GO classification.

Answer 1

[Jean-Karim Heriche]

As mentioned, you could group terms by semantic similarity. However, this may not help depending on your context. See for example some work I've been involved in here, the alternative approach is based on Finding New Order in Biological Functions from the Network Structure of Gene Annotations by Glass and Girvan. If the problem is that you have too many terms, you could also do your analysis with a slim ontology or only a branch of the ontology that is relevant to your topic of interest or a controlled vocabulary (i.e. a flat list of terms) instead of an ontology. Depending on what you're trying to do, simply picking a list of terms that cover processes you're interested in and covering everything else with an 'other' term may be the most efficient approach.

Answer 2

[Fratam]

I am not sure if this can help since I am not yet familiar with this analysis, but I recall I have read about GO semantic similarity analysis, and maybe upon further reading, you could find out if it suits your needs:

https://yulab-smu.top/biomedical-knowledge-mining-book/GOSemSim.html

As for my general knowledge about GO, I think you can find the common ancestor term of your terms and group by that.

[Jean-Karim Heriche]

Using an ancestor has the potential disadvantage that ancestor terms are less specific than child terms and that the same level of semantic specificity can be found at different levels on different branches. Usually one is interested in some level of precision in function description.

Answer 3

[geneontologyhelp]

Have you looked into using a Slim/subset? We have some information at http://geneontology.org/docs/go-subset-guide/ and you can use https://go.princeton.edu/cgi-bin/GOTermMapper with nearly zero prep, or you can use OWL2Tools. If this seems to be what you're looking for but the generic subsets (http://geneontology.org/docs/download-ontology/#subsets) aren't enough, we can walk you through creating a custom subset, although the need for a custom one is rare.

Answer 4

[EagleEye]

http://revigo.irb.hr/