[BioC] Advice with edgeR

[Paul Leo]

I've just starting exploring edgeR. Gotten an error I am unsure of...

I'm following the how-to with no changes

> d
DGEList:
$data
  lane3 lane8 lane5 lane6
1     0     0  1054   480
2    32     0   470   744
3     0     0   675   419
4     0     0   510   734
5   107     0   180  1138
81702 more rows ...

$lib.size
[1] 6482601  808363 7217381 7081739

$group
[1] ISOTYPE ISOTYPE CASE     CASE    
Levels: ISOTYPE CASE

> alpha
EBList:
$alpha
[1] 3.485649e-09

$common.dispersion
[1] 2.548059

d contains no columns where all counts are zero but there are a lot of
tags....
> sum(apply(d$data,1,sum)==0)
[1] 0
running the following:


chosen.lanes<-c("lane5","lane6","lane3","lane8")
considered<-rownames(the.counts)[apply(the.counts[,chosen.lanes],1,sum)!
=0
d<-DGEList(data = the.counts[considered,chosen.lanes], group =
c("CASE","CASE","ISOTYPE","ISOTYPE"),lib.size =
total.counts[chosen.lanes,"count"])
alpha<-alpha.approxeb(d)
alpha
ms<-deDGE(d,alpha=alpha$alpha)

##ERROR
Calculating shrinkage overdispersion parameters.
[quantileAdjust] Iteration (dot=1000) 1 :Error in approx(c(0, a +
c(diff(a)/2, 0)), c(-0.5, 0:mx), xout = p[i,  : 
  approx(): attempted to interpolate NA values

The error does not occur if I set doPoisson (yet) and I have a case
where probably 1/2 of the "tags" are not differentially expressed.

Any insight would be very welcome.


Thanks
Paul



> sessionInfo()
R version 2.10.0 Under development (unstable) (2009-07-24 r48986) 
x86_64-unknown-linux-gnu 

locale:
 [1] LC_CTYPE=en_AU.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_AU.UTF-8        LC_COLLATE=en_AU.UTF-8    
 [5] LC_MONETARY=C              LC_MESSAGES=en_AU.UTF-8   
 [7] LC_PAPER=en_AU.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_AU.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods
base     

other attached packages:
[1] edgeR_1.3.3   Biobase_2.5.4

loaded via a namespace (and not attached):
[1] tcltk_2.10.0 tools_2.10.0