[BioC] Agi4x44Preprocess - error on read.AgilentFE fucntion

James W. MacDonald jmacdon at uw.edu
Thu Jan 23 19:51:17 CET 2014


Hi Amanda,

Please don't take conversations off-list (e.g., use Reply-All).

On 1/23/2014 1:21 PM, amandafassis at usp.br wrote:
> Thank you very much James!
>
> In fact I had already done a test using readmaimages from limma 
> package  that worked very well inputting the data. :
>
> >library(limma)
> > targets<-readTargets("targets.txt")
>
> >x<-read.maimages(targets, source="agilent", green.only=TRUE)
>
> But I couldn't go further on pre processing data wijt Agi4x44 because 
> I got another error:
>
>  Erro em BGandNorm(x, BGmethod = normexp, NORMmethod = quantile, 
> foreground = MedianSignal,  :
>   'input' must be a RGList
>
> It seems that this package doesn't deal with single channel arrays... 
> Do you know something about that?

I haven't used Agi4x44PreProcess for a while now. I am not that familiar 
with the Agilent 4x44 arrays - I have only used the Salmonid arrays 
designed by the Koop lab, which are single color, so maybe there are two 
color 4x44 arrays out there that I don't know about. But if you look at 
the code for read.AgilentFE and BGandNorm, you can see that the author 
has done a bunch of work to stuff the results into an RGList, and then 
work within that framework. In other words, Agi4x44PreProcess sort of 
pretends that these arrays are two color even if they are not, and has 
to do a bunch of extra work because there is only data in one channel.

That is weird to me, as there is the EList structure, which is designed 
specifically for single color spotted arrays, but maybe 
Agi4x44PreProcess precedes the addition of EList to limma. I really 
don't know. But anyway, it is much easier and (to me) safer to just use 
limma directly:

x <- read.maimages(targets, source = "agilent", green.only = TRUE)
x <- backgroundCorrect(x, method = "normexp")
x <- normalizeBetweenArrays(x, method = "quantile")

Then you can filter out probes that are just above background if you 
like, and go directly on to model fitting. There is a worked example for 
these arrays in the limma User's Guide, starting on page 110 that you 
might like to read.

Best,

Jim


>
>
> Amanda Freire de Assis, MSc
> Molecular Immunogenetics Group
> University of São Paulo
> Av. Bandeirantes, 3900
> 14049-900 - Ribeirão Preto - Brazil
> +55 16 36023246
>
> ----- James W. MacDonald <jmacdon at uw.edu> escreveu:
> > Hi Amanda,
> >
> > On 1/23/2014 12:04 PM, Amanda F A Riccardi [guest] wrote:
> > > Hello everybody!
> > >
> > > I'm using Agi4x44Preprocess package to deal with pre processing 
> Agilent single color 4x44 whole mouse genome array, which data were 
> obtained using feature extraction software.
> > > The problem is that I can't download my data with the function 
> read.AgilentFE, I got an error - showed below - and the script just 
> stop and, instead mine, the author's example data is imported... I 
> used read.maimages function of the package limma as an alternative, 
> but I really would like to know what is wrong using read.AglentFE.
> > > Regards,
> > > Amanda
> > >
> > > -- output of sessionInfo():
> > >
> > >> targets <- read.targets(infile="targets.txt")
> > >
> > > Target File
> > > FileName Treatment Gerep
> > > CT1 CT1.txt CT 1
> > > CT2 CT2.txt CT 1
> > > CT3 CT3.txt CT 1
> > > TT1 TT1.txt TT 2
> > > TT2 TT2.txt TT 2
> > > TT3 TT3.txt TT 2
> > > PT1 PT1.txt PT 3
> > > PT2 PT2.txt PT 3
> > > PT3 PT3.txt PT 3
> > >
> > >> dd <-read.AgilentFE(targets,makePLOT=TRUE)
> > > Read CT1.txt
> > > Read CT2.txt
> > > Read CT3.txt
> > > Read TT1.txt
> > > Read TT2.txt
> > > Read TT3.txt
> > > Read PT1.txt
> > > Read PT2.txt
> > > Read PT3.txt
> > > INPUT DATA DOES NOT CONTAIN - Sequence and chr_coord
> > > SCANN THE DATA USING AFE 9.5.3.1
> >
> > ^^^^^^^^^ That is supposed to be your hint. What it means is that the
> > sequence and chromosome coordinate data are not available, and that you
> > should use Agilent Feature Extraction version 9.5.3.1 (or higher, I
> > presume).
> >
> > If you don't have these data available, read.AgilentFE() will fail, as
> > this test is hard-coded into the function. In other words, unless you
> > upgrade your AFE version, you will have to use read.maimages() directly.
> >
> > Best,
> >
> > Jim
> >
> >
> > > Erro em read.AgilentFE(targets, makePLOT = TRUE) :
> > > the script will stop now
> > >
> > >
> > > --
> > > Sent via the guest posting facility at bioconductor.org.
> > >
> > > _______________________________________________
> > > Bioconductor mailing list
> > > Bioconductor at r-project.org
> > > https://stat.ethz.ch/mailman/listinfo/bioconductor
> > > Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor
> >
> > --
> > James W. MacDonald, M.S.
> > Biostatistician
> > University of Washington
> > Environmental and Occupational Health Sciences
> > 4225 Roosevelt Way NE, # 100
> > Seattle WA 98105-6099
> >

-- 
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099



More information about the Bioconductor mailing list