[BioC] affypara

Valerie Obenchain vobencha at fhcrc.org
Wed Oct 13 21:42:14 CEST 2010


On 10/13/2010 04:25 AM, danieladna wrote:
> On 10/08/2010 09:14 AM, danieladna wrote:
>   
>> Hello,
>> I have been running the affy package for a while now without running into any 
>> problems.
>> Now I would like to use affyPara as I have 600 .CEL files to analyze.
>> All my packages are up to date and all cdf files are installed.
>> Unfortunately affyPara cannot load any cdf files. I am not sure why. Does 
>> anyone 
>>
>> have an idea how I can fix this problem?
>> The failure message I get is following:
>> Error in checkForRemoteErrors(lapply(cl, recvResult)) : 
>>   31 nodes produced errors; first error: Could not obtain CDF environment, 
>> problems encountered:
>> Specified
>>     
>  environment does not contain HuEx-1_0-st-v2
>   
>> Library - package huex10stv2cdf not installed
>> Bioconductor - could not connect
>> Calls: rmaPara -> preproPara -> clusterCall -> checkForRemoteErrors
>> Execution halted
>>
>> I am positive that huex10stv2cdf is installed, as I have been running it with 
>> affy successfully from the same location.
>> Many thanks for your help and ideas,
>> Daniela
>>
>>
>>   
>>     
> Hi Daniela,
>
> Seeing more of your code would be helpful for troubleshooting the
> problem.  Please provide a chunk of code that describes how you are
> inputting the cdf files and what you are distributing to the nodes. 
>
> Also include the output of sessionInfo() from one of the nodes.
> For example, if cl is your cluster then
> clusterEvalQ(cl, sessionInfo())
>
> I am
>  also ccing the package maintainer on this response.
>
> Valerie
>
>
> Hi Valerie,
> Many thanks for your reply.
> My code looks like this:
>
> library(affyPara)
> library(snow)
> library(huex10stv2hsensecdf)
>
> cl<-makeCluster()
> clusterEvalQ(cl, sessionInfo())
>
> path<-"/home/daniela/test"
> files <- list.celfiles(path,full.names = T)
> distList<-distributeFiles(files,protocol="SCP")
> eset<-rmaPara(distList$files, cdfname="huex10stv2hsensecdf ",cluster=cl)
> sessionInfo()
> e<-exprs(eset)
>
> write.exprs(file="samples.txt")
>
>
> removeDistributedFiles("/home/daniela/test")
> stopCluster(cl)
>
> The session info like this: and is repeated for all processors I have been 
> requesting.
> [[3]]
> 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 datasets  utils     methods   base     
>
> other attached packages:
> [1] snow_0.3-3 Rmpi_0.5-8
>
> And this is the error message I get:
>
> There were 24 warnings (use warnings() to see them)
> Error in .getObjectType(object) : Object has an unknown type
> Calls: rmaPara -> preproPara -> .getObjectType
> Execution halted
>
> Do I really need to manually distribute the files with the distributeFiles 
> command? Or could I leave this command out since
> I am working on a cluster and all the files are already on all the available 
> nodes because our shared file system.
> I think that that distributeFiles command might cause following error: 
> Permission denied (publickey,keyboard-interactive).
> lost connection. Is this a possible cause for my problems?
>   
If you are computing on a single machine with many processors (shared
memory) then you don't need distributeFiles. The publickey error may be
related to the use of distributeFiles but the CDF environment error is
most likely not.

Problems related to the CDF error (in a non-parallel environment) have
been discussed on the mailing list which may or may not be helpful,

    https://stat.ethz.ch/pipermail/bioconductor/2010-April/032925.html

and

    https://stat.ethz.ch/pipermail/bioconductor/2010-April/032952.html


Please send the results of

    library(affy)
    library(huex10stv2hsensecdf) 
    affyBatch <- ReadAffy("f1",  cdfname="huex10stv2hsensecdf " )
    res <- rma(affyBatch)

and

    ff <- just.rma(f1, cdfname="huex10stv2hsensecdf " )

where f1 is an absolute path to a single cel file.  Also, please send
the huex10stv2hsensecdf package so I can do some more testing.


Valerie




> I very much appreciate 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
>



More information about the Bioconductor mailing list