GOterm semantic similarity with non-reference organism
1
0
Entering edit mode
@cpcantalapiedra-7532
Last seen 9.5 years ago
Spain

Is there any way to reduce the redundancy of a set of GO terms when working with a non-reference organism? I mean something similar to REVIGO. The problem with the last is that has a limit in the number of GO terms that can be submitted.

I have read the documentation of GoSemSim and it seems that can work only with supported organisms (though I expected that maybe it would have a function like buildGOmap from clusterProfiler). I have tried to work with goProfiles too, but there is no way to make it work without a orgPackage option (though in the examples from the vignettes it is not used nor mentioned).

edit: I finally was able to make goProfiles work with the basicProfile() parameter idType = "GOTermsFrame". This way, it ignores the orgPackage and anotPackage parameters since you provide the GeneID <--> GOID mappings goProfiles needs.

Any advice?

Thank you in advance

ps.:  I have already done enrichment analysis (with topGO) and it is not what I mean in this post.

gosemsim go gene ontology • 2.1k views
ADD COMMENT
0
Entering edit mode

GOSemSim do support non-reference organisms. The organism parameter only used for IC-based methods those depend on information content data within GOSemSim packages. These IC data are species specific and pre-calculated. I may add supports for calculating IC in real time and support non-reference organisms in future.

 

For Wang's measure, it used the structure of the whole GO DAG graph and organism parameter will be omitted in goSim and mgoSim function. For gene-based calculation, we use organism parameter to mapping gene to GO.

You want to remove redundant of enrichment result, there is no organism restriction if you use Wang's measure implemented in GOSemSim.

 

Bests,

Guangchuang

 

ADD REPLY
0
Entering edit mode

For your information, I will implement a function, simplify, to remove redundant GO terms in enrichment result within the clusterProfiler package.

ADD REPLY
0
Entering edit mode
@kirsten-gotting-7390
Last seen 9.5 years ago
Kansas City, Missouri

For what purpose do you need a smaller set?  

I've tried (somewhat unsuccessfully) using GOSim, it's similar to GoSemSim but it allows you to put in the annotation of GO terms you are using, if using and annotation from a non-model organism: http://www.bioconductor.org/packages/release/bioc/html/GOSim.html

Another tool that is outside of Bioconductor is GO trimming : http://lucy.ceh.uvic.ca/go_trimming/cbr_go_trimming.py

I've had some success with this tool, but it didn't reduce as much as I would have liked (I started with around 1500 GO terms and it reduced the set to around 1200).

Also, you could consider mapping your GO terms to the one of the GOslim sets, this will give you a set of terms that is much more generalized: http://geneontology.org/page/go-slim-and-subset-guide

ADD COMMENT
0
Entering edit mode

thank you very much Kirsten. I will try those other alternatives.

ADD REPLY

Login before adding your answer.

Traffic: 587 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6