[BioC] Limma and Quantarray files

Gordon K Smyth smyth at wehi.EDU.AU
Wed Oct 4 13:56:03 CEST 2006


Dear Alessio and Gavin,

Thanks for pointing out the problem with reading quantarray files.  The bug has now been fixed in
limma 2.9.1.  limma now reads the annotation columns for quantarray files as well.

Best wishes
Gordon

> Date: Wed, 4 Oct 2006 09:43:50 +0100
> From: "Gavin Kelly" <gavin.kelly at cancer.org.uk>
> Subject: Re: [BioC] Limma and Quantarray files
> To: <bioconductor at stat.math.ethz.ch>
> Message-ID:
> 	<1BFC6DAE48E6D4478BDFF41C012CD311FC4FB4 at uk-lif-lexc02.crwin.crnet.org>
> Content-Type: text/plain;	charset="us-ascii"
>
>> Hi,
>>
>> I've a problem using limma with quantarray output files. I've changed
> all output files to ANSI format,
>> because the original Unicode format is not accepted by limma.
>> I'm able to use one file per time without any trouble but when I try
> to load more than
>> one file using a "target" file using,
>>
>>
>>target<-readTargets(file="targets.txt",path="C:\\Playground\\CrossTest"
> )
>>
>>
>>RG<-read.maimages(target,source="quantarray",path="C:\\Playground\\Cros
> sTest")
>>
>> I receive an error message like:
>>
>> Read C:\Playground\CrossTest/PA52-Export.txt
>> Error in if (skip > 0) readLines(file, skip) :
>>         argument is of length zero
>>
>> How can I solve this problem?
>>
>> Kind Regards
>>
>> Alessio
>>
>
> Alessio,
> I had a similar problem in a recent version of limma (not sure which
> version you're using).  If you edit the source for read.maimages, and
> look for the second instance of the line:
>
> skip <- grep("Begin Data", firstfield)
>
> and comment this out.  Save this as your own function, and it should
> work okay.  This will only work if all your quantarray files have the
> data (as flagged by the "Begin Data" string) starting at the same line.
> If this is not the case, then you'll need to find the instance of the
> phrase "nlines=100" just above the skip assignment, and remove
> it/customise it to your files (and uncomment the skip assignment)
>
> Hope this helps - Gavin
>
> --
> Gavin Kelly (DPhil) Senior Statistician
> Bioinformatics & Biostatistics, Room 708
> Cancer Research UK
> 44 Lincoln's Inn Fields
> London WC2A 3PX
> United Kingdom



More information about the Bioconductor mailing list