Visualize GO term relationships with only ID

Question

[grapestory]

Hi Biostars!

I am new here so please let me know if I am doing anything incorrectly. My question is: How can I visualize the relationship between GO terms using just the GO ID (ex. GO:0008152)? My data does not come from enrichment analysis so I do not have any information about gene expression or p-values. I just have a list of GO IDs. Ideally I would like to use R to visualize this but I am open to anything.

From searching on this site for similar questions, most people had other metadata that programs like REVIGO could use. I was unable to find any suggestions for building graphs using just GO IDs. The AmiGO website (http://amigo.geneontology.org/visualize?mode=client_amigo) does have a web interface that does this, however it breaks with more than 100-200 terms. My data has 300-1000 terms.

Thank you in advance for the help,

grapestory

Answer 1

[jared.andrews07]

ontoProc can do this very easily. See the onto_plot2 function. My guess is that anything you make will be an incomprehensible hairball mess with that many terms, but you can give it a shot.

[grapestory]

Well, this did do exactly what I was hoping for. Thank you, Jared! It did also make a giant hairball mess like you predicted. Do you have any suggestions for a way to more intelligently visualize this data? I am really just trying to figure out what "higher" GO terms most of my list falls under?

[jared.andrews07]

Hmm, it seems like the ontologyIndex and ontologyPlot packages might be a better fit then. They let you find ancestor terms and also remove the intermediary links and/or descendants if you want.

There are simple examples of both in the ontologyPlot vignette, which seems more like what you want. It may still be a mess, but you might be able to highlight your common hubs.