[BioC] affyPara

Valerie Obenchain vobencha at fhcrc.org
Thu Oct 21 21:15:47 CEST 2010 

Resending this message under the appropriate name for tracking purposes.


Hi Daniela,

Thank you for sending the output and cdf file. It looks as though
affyPara has  bug in rmaPara when a user supplies a cdf that does not
match the cdf information encoded in the cel files. The problem appears
to be with the .initAffyBatchSF function that does initialization of the
AffyBatch object on the workers. This function  does not allow the
passing of a user supplied cdf file so the cdf used is the default taken
from the cel files.

The cel files you are using are looking for file HuEx-1_0-st-v2 which is
apparently in package huex10stv2cdf. The workers rmaPara  spawns can't
find this package because it isn't loaded (instead we have
huex10stv2hsensecdf loaded). The workers also aren't aware that you want
to use the huex10stv2hsensecdf package because .initAffyBatchSF does not
pass the cdfname="huex10stv2hsensecdf" information to them. I will
contact the package maintainer about this problem and cc you on the message.

An interim work around would be to use just.rma on the workers. Your
result will be a list of esets (one from each worker).

    library(affyPara)
    library(huex10stv2hsensecdf)
    setwd("/home/daniela/test")
    cl <- makeCluster()
    # load affy on each worker so they have access to just.rma
    clusterEvalQ(cl, library(affy))
    eset <- clusterEvalQ(cl,
                 just.rma(list.celfiles(), cdfname="huex10stv2hsensecdf"))
    stopCluster(cl)

Hopefully this approach allows you to get your analysis done.

Valerie


On 10/16/2010 09:19 AM, Daniela Kamir wrote:

> > Hi Valerie,
> > This is Daniela (danieladna) from the Bioconductor forum. I posted for
> > help about the package affyPara.
> > You asked me to send you my results.
> > These are my output results:
> >
> >
>   
>> > > library(affydata)
>> > >  celpath <- system.file("celfiles", package="affydata")
>> > > fns <- list.celfiles(path=celpath,full.names=TRUE)
>>     
> >      
>   
>> > > abatch<-ReadAffy(filenames="1031.CEL", cdfname="huex10stv2hsensecdf")
>> > > abatch
>>     
> >
> > Attaching package: 'huex10stv2hsensecdf'
> >
> > The following object(s) are masked from 'package:hgu133acdf':
> >
> >     i2xy, xy2i
> >
> > AffyBatch object
> > size of arrays=2560x2560 features (16 kb)
> > cdf=huex10stv2hsensecdf (304497 affyids)
> > number of samples=1
> > number of genes=304497
> > annotation=huex10stv2hsensecdf
> > notes=
> >
>   
>> > > res<-rma(abatch)
>>     
> > Background correcting
> > Normalizing
> > Calculating Expression
>   
>> > > res
>>     
> > ExpressionSet (storageMode: lockedEnvironment)
> > assayData: 304497 features, 1 samples
> >   element names: exprs
> > protocolData
> >   sampleNames: 1031.CEL
> >   varLabels and varMetadata description:
> >     ScanDate: NA
> > phenoData
> >   sampleNames: 1031.CEL
> >   varLabels and varMetadata description:
> >     sample: arbitrary numbering
> > featureData: none
> > experimentData: use 'experimentData(object)'
> > Annotation: huex10stv2hsensecdf
> >
> > But I do not manage to run just.rma:
> >
>   
>> > > ff<-just.rma(~/Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST,cdfname="huex10stv2hsensecdf")
>>     
> > Error: unexpected '/' in "ff<-just.rma(~/"
>   
>> > > ff<-just.rma(/Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST,cdfname="huex10stv2hsensecdf")
>>     
> > Error: unexpected '/' in "ff<-just.rma(/"
>   
>> > > ff<-just.rma(Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST,cdfname="huex10stv2hsensecdf")
>>     
> > Error: unexpected input in "ff<-just.rma(Expression-Exon/LBL__HuEx-1_"
>   
>> > > ?just.rma()
>> > > ff<-just.rma(abatch,cdfname="huex10stv2hsensecdf")
>>     
> > Error: file names must be specified using a character vector, not a ‘list’
>   
>> > > getwd()
>>     
> > [1]
> > "/home/danielak/Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST"
>   
>> > > ff<-just.rma(home/danielak/Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST,cdfname="huex10stv2hsensecdf")
>>     
> > Error: unexpected input in
> > "ff<-just.rma(home/danielak/Expression-Exon/LBL__HuEx-1_"
>   
>> > > ff<-just.rma(/home/danielak/Expression-Exon/LBL__HuEx-1_0-st-v2/Level_1/CONTROLS/TEST,cdfname="huex10stv2hsensecdf")
>>     
> > Error: unexpected '/' in "ff<-just.rma(/"
> >
>   
>> > > sessionInfo()
>>     
> > R version 2.11.1 (2010-05-31)
> > x86_64-unknown-linux-gnu
> >
> > locale:
> >  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
> >  [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
> >  [5] LC_MONETARY=C              LC_MESSAGES=en_US.UTF-8   
> >  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
> >  [9] LC_ADDRESS=C               LC_TELEPHONE=C            
> > [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
> >
> > attached base packages:
> > [1] stats     graphics  grDevices utils     datasets  methods   base     
> >
> > other attached packages:
> > [1] huex10stv2hsensecdf_12.1.0 hgu133acdf_2.6.0          
> > [3] affydata_1.11.10           affy_1.26.1               
> > [5] Biobase_2.8.0             
> >
> > loaded via a namespace (and not attached):
> > [1] affyio_1.16.0         preprocessCore_1.10.0 tools_2.11.1     
> >
> > Please find attached the cdf file you asked me for. I really hope you
> > can help me find the reason why affyPara won’t run or rather won’t
> > detect the cdf file.
> > I really appreciate your help,
> > Daniela 
>   









On 10/15/2010 04:12 PM, Valerie Obenchain wrote:
> Hi Daniela,
>
> On 10/14/2010 04:17 PM, danieladna wrote:
>   
>> Hi Valerie,
>> Unfortunately ReadAffy doesn't manage to load one file only.
>>   
>>     
> ReadAffy can read a single file at a time. See the examples in ?ReadAffy
> then try
>
>     require(affydata)
>     celpath <- system.file("celfiles", package="affydata")
>     fns <- list.celfiles(path=celpath,full.names=TRUE)
>     # read a single file
>     abatch <- ReadAffy(filenames=fns[2])
>     # alternatively, read all files in a directory
>     abatch <- ReadAffy(celfile.path=celpath)
>
>   
>> I do manage to load all CEL files that are in the current location though.
>> Therefore I use:
>> data<-ReadAffy()
>> data at cdfName<-"cdfname"
>> eset<-rma(data)
>>   
>>     
> It is ok if you want to use all files or a single file for this example;
> a single file would simply be faster. I am interested in the output of
> the example because I want to confirm that no CDF errors are thrown when
> reading a cel file with the same huex10stv2hsensecdf package you are
> using in your affyPara code. I'd like to help further with this problem
> but need a couple of things from you in order to do so.  Please
>
> (1) run this code and provide the output
>
>     library(affy)
>     library(huex10stv2hsensecdf)
>     data <- ReadAffy()
>     data at cdfName <- "huex10stv2hsensecdf"
>     est <- rma(data)
>
>
> (2) send me the huex10stv2hsensecdf package you are using
>
> Thank you,
> Valerie
>
>
>   
>> I get most of my CDF files from Brainarray.
>> Thanks for your help,
>> Daniela
>>
>>
>>
>>       
>> 	[[alternative HTML version deleted]]
>>
>> _______________________________________________
>> 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
>>
>>     
> _______________________________________________
> 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
>

More information about the Bioconductor mailing list