[BioC] problems with gcrma just.gcrma

[Elisabetta Manduchi]

Hi Seth,
thanks for the futher suggestions.
In response, first, the output of sessionInfo() is:

---
R version 2.4.0 (2006-10-03)
i386-pc-mingw32

locale:
LC_COLLATE=English_United States.1252;LC_CTYPE=English_United 
States.1252;LC_MONETARY=English_United 
States.1252;LC_NUMERIC=C;LC_TIME=English_United States.1252

attached base packages:
[1] "splines"   "tools"     "methods"   "stats"     "graphics" 
"grDevices"
[7] "utils"     "datasets"  "base"

other attached packages:
mgu74av2probe   mgu74av2cdf         gcrma   matchprobes          affy
      "1.14.0"      "1.14.0"       "2.6.0"       "1.6.0"      "1.12.2"
        affyio       Biobase
       "1.2.0"      "1.12.2"
---

I had done a fresh install of the most recent R version (2.4.0) and of the 
bioconductor packages (through getBioC.R) just before running gcrma.
Aren't the versions for affy and affyio above the most recent ones?
Second, I've followed your subsequent instruction to enter the debugger 
and here is the copy-paste of the session:

---
> library(affy)
> debug(read.probematrix)
> 
data.gcrma<-just.gcrma(filenames=cel.files,type=c("fullmodel"),normalize=T,optimize.by=c("memory"),fast=T,samepleNames=sample.names)
Computing affinities.Done.
debugging in: read.probematrix(filenames = filenames, which = "pm", 
cdfname = cdfname)
debug: {
     auxnames <- as.list(substitute(list(...)))[-1]
     filenames <- .Primitive("c")(filenames, auxnames)
     which <- match.arg(which, c("pm", "mm", "both"))
     if (verbose)
         cat(1, "reading", filenames[[1]], "to get header information")
     headdetails <- .Call("ReadHeader", filenames[[1]], PACKAGE = "affyio")
     dim.intensity <- headdetails[[2]]
     ref.cdfName <- headdetails[[1]]
     if (is.null(cdfname))
         Data <- new("AffyBatch", cdfName = ref.cdfName, annotation = 
cleancdfname(ref.cdfName,
             addcdf = FALSE))
     else Data <- new("AffyBatch", cdfName = cdfname, annotation = 
cleancdfname(ref.cdfName,
         addcdf = FALSE))
     cdfInfo <- as.list(getCdfInfo(Data))
     cdfInfo <- cdfInfo[order(names(cdfInfo))]
     .Call("read_probeintensities", filenames, rm.mask, rm.outliers,
         rm.extra, ref.cdfName, dim.intensity, verbose, cdfInfo,
         which, PACKAGE = "affyio")
}
Browse[1]> n
debug: auxnames <- as.list(substitute(list(...)))[-1]
Browse[1]> n
debug: filenames <- .Primitive("c")(filenames, auxnames)
Browse[1]> n
debug: which <- match.arg(which, c("pm", "mm", "both"))
Browse[1]> n
debug: if (verbose) cat(1, "reading", filenames[[1]], "to get header 
information")
Browse[1]> n
debug: headdetails <- .Call("ReadHeader", filenames[[1]], PACKAGE = 
"affyio")
Browse[1]> n
debug: dim.intensity <- headdetails[[2]]
Browse[1]> n
debug: ref.cdfName <- headdetails[[1]]
Browse[1]> n
debug: if (is.null(cdfname)) Data <- new("AffyBatch", cdfName = 
ref.cdfName,
     annotation = cleancdfname(ref.cdfName, addcdf = FALSE)) else Data <- 
new("AffyBatch",
     cdfName = cdfname, annotation = cleancdfname(ref.cdfName,
         addcdf = FALSE))
Browse[1]> n
debug: cdfInfo <- as.list(getCdfInfo(Data))
Browse[1]> n
debug: cdfInfo <- cdfInfo[order(names(cdfInfo))]
Browse[1]> n
debug: .Call("read_probeintensities", filenames, rm.mask, rm.outliers,
     rm.extra, ref.cdfName, dim.intensity, verbose, cdfInfo, which,
     PACKAGE = "affyio")
Browse[1]> n
Error in read.probematrix(filenames = filenames, which = "pm", cdfname = 
cdfname) :
         VECTOR_ELT() can only be applied to a 'list', not a 'symbol'
---

Do you think this might be indeed a bug with one of the libraries?
Thanks again,
Elisabetta

---

On Mon, 4 Dec 2006, Seth Falcon wrote:

> Elisabetta Manduchi <manduchi at pcbi.upenn.edu> writes:
>
>> Seth,
>> thanks for your prompt feedback. I'll be able to provide the
>> sessionInfo output tomorrow from my office, since I was running this
>> locally from my desktop there.
>> I did however run my code after a fresh install of R 2.4.0 and of
>> getBioC.R.
>> I had another error with my previous version or R and bioconductor and
>> I had pinned that down to a problem with the probeset file for that
>> version (1.12.0), which appeared to be damaged. This prompted me to do
>> a fresh install, so I imagine that the package versions should have
>> all been in sync. But I'll in any case check on what you suggest.
>
> I have to admit that I'm sort of hoping that you discover that your
> affy and affyio packages are somehow not up-to-date.  That would be
> the easiest thing to fix :-)
>
> If everything turns out to match, then it suggests there is a bug in
> affy or affyio.  Perhaps you could try to gather some more info to
> post to the list (please, not to me personally) by doing something
> like the following:
>
> library(affy)
> debug(read.probematrix)
>
> # run your code that causes the error
> # when the code calls read.probematrix you will enter the debugger.
> ## hit the 'n' key and then <enter> to execute the next line of code.
> ## Just keep doing that until the error occurs and then send the
> ## copy/paste of the session.
>
> + seth
>
> -- 
> Seth Falcon | Computational Biology | Fred Hutchinson Cancer Research Center
> http://bioconductor.org
>