[BioC] makeGOGraph internode distance

[Loren Engrav]

Continue to use MakeGOGraph and GOGraph and Illustrator for GO tree with 350
nodes

and the problem of overlapping labels seems solved with

postscript("file.ps", width = 100.0, height = 100.0, paper="special"); plot
(xctree); dev.off()

as suggested by thatsanicehatyouhave at mac.com, thank you

This makes the system of makeGOGraph -> GOGraph -> Illustrator/Acrobat even
better for large trees


> From: Loren Engrav <engrav at u.washington.edu>
> Date: Mon, 19 Nov 2007 19:50:53 -0800
> To: <bioconductor at stat.math.ethz.ch>
> Conversation: [BioC] makeGOGraph internode distance
> Subject: Re: [BioC] makeGOGraph internode distance
> 
> Thank you
> 
> I am learning two things, how to use R/Bioconductor and how to do the email
> Bear with me please, I will learn both
> 
> David Bodanis in E=mc^2 when discussing geezers doing research said, geezers
> may have good ideas, that is not the problem, their problem is that they are
> afraid to take risk, entering the R/Bioconductor email fray demonstrates
> that I can take risk
> 
> And I may have a work around that is better than any solution
> 
> Save graph as postscript
> Open in Illustrator
> Scale 400%
> Rotate labels 45 degrees
> Set font to 6point
> And voila, a nice big induced GO tree
> Never mind if I will ever understand the tree, it is the process, not the
> result
> 
> But as for the problem
> 
>> you probably have more nodes than you can easily render on the available
> plotting device
> 
> True
> length(go2tree.leaves) returned 428 nodes
> 
>> You could make it larger
> 
> I have tried this, but failed, am sorry
> Like I tried
>> par (pin = c(100,100))
>> plot (xctree)
> It say Error in plot.new() : plot region too large
> So I try
>> par (pin = c(50,50))
>> plot (xctree)
> It say Error in plot.new() : plot region too large
> Etc, etc, etc
> Till I choose > par (pin = c(10,10))
> Which is pretty much where I started
> So I clearly do not know how to make it larger, am sorry
> 
>> how would you know that the tree has changed?
> Open them in Illustrator at 400% scale and I cannot see significant
> difference
> 
>> You should read the Rgraphviz vignette
> I have read that so much that I have almost memorized it, but I could not
> make it solve this problem, am sorry
> 
>> You may find that you would rather use graphviz
>> You could investigate the use of the imageMap function (geneplotter)
> 
> True but, I think the combo of makeGOGraph (from genes) and xctree.leaves
> [1:428]  and GOGraph (from leaves trimmed of junk) and Illustrator is
> extremely powerful to make a big GO tree, is in fact pawerful enough, again
> never mind if I ever understand the biology, it is process not result
> 
> I see you wrote both, cool, you might add to the documentation something
> about large trees and to finish the study for LARGE trees the user might do
> makeGOGraph, get the nodes, cull the junk, plot them up with GOGraph, and
> load them up into Illustrator to wrap it up
> 
> which I will now do for MF
> 
> Thank you
> 
> Loren Engrav, MD
> Univ Washington
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives:
> http://news.gmane.org/gmane.science.biology.informatics.conductor