[BioC] affylmGUI gene list from Venn diagram, was: (no subject)

[James Wettenhall]

Simon,

Please check your steps one and a time and find out where 
the '100' comes from.  I would suspect that you have 
accidentally defined tstat from the ?vennCounts example in 
limma instead of using:
tstat <- ContrastParameterizationList[[1]]$eb$t
as described in affylmGUI/doc/extract.pdf.

Here's an example where I run classifyTestsF 
on a file "EstrogenContrastsComputed.lma" which I have saved 
from affylmGUI:

> library(affy)  # for getCdfInfo method
> load("EstrogenContrastsComputed.lma")
> tstat <- ContrastParameterizationList[[1]]$eb$t
> nrow(tstat)
[1] 12625
> clas <- classifyTestsF(tstat,df=10,p.value=0.05)
> nrow(clas)
[1] 12625

> clas[1:10,] # Just look at the first ten rows.
        Contrast1 Contrast2 Contrast3
   [1,]         0         0         0
   [2,]         0         0         0
   [3,]         0         0         0
   [4,]         0         0         0
   [5,]         0         0         0
   [6,]         0         0         0
   [7,]        -1        -1        -1
   [8,]         0         0         0
   [9,]         0         0         0
  [10,]         0         0         0

We see that the 7th gene (out of 12625) is down-regulated in all 
three contrasts.  To find out which gene that is look at 
geneIDs, e.g. defined from getCdfInfo(RawAffyData) or use
data.frame(geneIDs,clas)
to combine the geneIDs with the classification.

Hope this helps,
James


 On Sat, 19 Jun 2004, Simon Kidd wrote:
> >Sorry, my mistake.  What you need to use in the data.frame is
> >the 'clas' object from classifyTestsF, not the 'a' object
> >from vennCounts.
> >
> >clas <- classifyTestsF(tstat,df=10,p.value=0.05)
> >class(clas) <- "matrix"
> >data.frame(geneIDs=geneIDs,clas)
> 
> With this I get a slightly different error message, here is what I 
> typed and the error message:
> --------------
> load("rma.lma")
> tstat <- ContrastParameterizationList[[1]]$eb$t
> tstat <- matrix(rt(300,df=10),100,3)
> tstat[1:33,] <- tstat[1:33,]+2
> clas <- classifyTestsF(tstat,df=10,p.value=0.05)
> a <- vennCounts(clas)
> geneIDs <- ls(getCdfInfo(RawAffyData))
> class(clas) <- "matrix"
> data.frame(geneIDs=geneIDs,clas)
> 
> Error in data.frame(geneIDs = geneIDs, clas) :	arguments imply 
> differing number of rows: 14010, 100
> 
> -------------
> 
> I dont know if it matters but I have eight targets of which only four 
> are used for contrasts, will that make a difference?
> 
> thanks for your help.
> 
> Simon
> 
> 
>