[BioC] Limma_Agilent_2color_Array

[Wolfgang Huber]

Dear Abhilash

> As you directed, I tried arrayQualityMetrics. I used feature extracted txt
> file  from Agilent  arrays and created RGList. It has given the following
> error, I am wondering whether it is  because I didnot try to create
> assayData, featureData and phenoData.
> arrayQualityMetrics(expressionset = array1,
> + outdir = "array")
> *The directory 'array' has been created.
> Error in seq_len(ncol(x)) : argument must be non-negative
> 
> *What could I do to solve this problem?

First of all, give the output of sessionInfo().
Second, you could tell us a bit more about what "array1" is.

 Best wishes
	Wolfgang

> Regards,
> Abhilash
> 
> 
> On Wed, Apr 16, 2008 at 11:07 PM, Wolfgang Huber <huber at ebi.ac.uk> wrote:
> 
>> Dear Abilash,
>>
>> you could try to run the arrayQualityMetrics report from the eponymous
>> package (and preferably a current devel version [1]). The plots in there
>> would help you to assess both whether there are gross problems in the
>> data, and how much and what type of normalisation is needed.
>>
>> Adequate normalisation removes unwanted technical variation while it
>> maintains the interesting biological signal in the data. The best way to
>> assess whether this is goal is achieved is by looking at the behaviour
>> of the controls that were done as part of the experiment.
>>
>> [1]
>>
>> http://www.bioconductor.org/packages/2.2/bioc/html/arrayQualityMetrics.html
>>
>> Best wishes
>>  Wolfgang
>>
>> ------------------------------------------------------------------
>> Wolfgang Huber  EBI/EMBL  Cambridge UK  http://www.ebi.ac.uk/huber
>>
>>
>> 16/04/2008 18:02 Abhilash Venu a écrit
>>> Hi all,
>>>
>>> I have a few questions. I am analyzing  20 experiments  performed  in
>>> Agilent 4x44K. Tumor and adjacent normal were labeled with Cy5 and Cy3
>>> respectively. The experiments include two dye swaps also. The feature
>>> extraction were performed using Agilent feature extraction software
>> version
>>> 8.5. The txt files obtained after feature extraction, subjected to
>> further
>>> analysis by limma. The script has given below
>>> txt_files <- dir(pattern=".txt")
>>> RG<-read.maimages(txt_files, source="agilent")
>>> design<-c(1,1,1,1,1,1,1,1,1,1,1,-1,1,1,1,1,1,1,1,-1) # this include two
>> dye
>>> swap also
>>> Rgene<-backgroundCorrect(RG,method="normexp")
>>> fit<-lmFit(Rgene,design)
>>> fit<-eBayes(fit)
>>> topTable(fit,adjust="fdr", number=100)
>>>
>>> As a result I get the format in which 'B' indicates what parameter? And
>> is
>>> this analysis method is enough to get a good data?
>>> In Agilent arrays, I think generally it is normalized, in that case how
>> the
>>> data will be affected by normalizing it again using above method?
>>> How can we decide which normalization methods should be used? Do you
>> think
>>> its always better to decide by plotting different graphs?
>>>
>>> Comments and answers will be appreciated.
>>>
>>> Regards,
>>> Abhilash
>>>
>>>
> 
> 
>