The Best Visualization For Go Term Linkage

Question

[k.nirmalraman]

The following is basically a loud brainstorming.

In order to visualize the GO TERMS I have identified to be enriched in my sample, I have the following two questions.

1. I have a long list of GO terms and would like to visualize them based on their linkage (say in R) i.e., based on the genes that are shared among the GO pathway gene sets and not based on my data set.

2. Is it possible to shrink these GO terms based on some ancestor in GO topology (any tools?), so I do have all the similar groups are shrinked and so direct interpretation can be obtained visually.

(say if I have 10 GO terms that concern Mitochondria, I would like them to be grouped together, in a network). (ReviGO does this quite well), but any other suggestions or customizable source codes will be great!

Additional Question:

What is the usual approach one would take for pathway enrichment for multiple samples/conditions.

GSEA or other enrichment methods are limited to pairwise comparisons... (which is of course is the funda of enrichment).

What I would like to do..... is identify the pathways that are particularly enriched under each conditions in comparison to the rest.

(GSVA is a Bioconductor R package which provides one such enrichment approach, but I would like other options to validate this)

Thanks!

[Ben]

Not sure if you're asking for recommendations re: GO slimmers or enrichment visualisations (or both); for the latter, if it's a model organism I think GOrilla is pretty clear

Answer 1

[Woa]

For your First question: Enrichment Map can possibly do it with some tweaks

http://www.baderlab.org/Software/EnrichmentMap

and check this reference, particularly Fig. 2

http://www.biomedcentral.com/content/pdf/1756-0500-3-10.pdf