[BioC] blue fuse output and limma

[Gordon Smyth]

At 04:34 AM 4/08/2004, marchiem\@libero\.it wrote:
>Hello everybody,
>we are evaluating a new image analysis software called Blue Fuse that uses 
>a Bayesian approach to estimate the background. The outputs per each 
>channell of this software are values with the background already subtracted.
>How can I read in these output in BioC, specifically in limma?
>
>I've used read.maimages and I've specified the columns just for foreground 
>signal and I put NULL for the background:
>RGblue<-read.maimages(files3,columns=list(Rf="AMPCH2",Gf="AMPCH1",Rb=NULL,Gb=NULL))

The recommended way to read BlueFuse data into limma is

RG <- 
read.maimages(files,columns=list(Rf="AMPCH2",Gf="AMPCH1",Rb="AMPCH2",Gb="AMPCH1"))
RG <- backgroundCorrect(RG, method="none")

This question has been previously addressed on this mail list, see:
https://www.stat.math.ethz.ch/pipermail/bioconductor/2004-February/003674.html

BlueFuse also has a confidence value column, although I am not sure of the 
precise name of this column. (There isn't any online documentation.) To 
make use of this column, you could include a 'wt.fun' argument to 
read.maimages(), for example:

mywtfun <- function(x) x[,"CONFIDENCE"]
RG <- 
read.maimages(files,columns=list(Rf="AMPCH2",Gf="AMPCH1",Rb="AMPCH2",Gb="AMPCH1"),wt.fun=mywtfun)

where "CONFIDENCE" is the name of the confidence value column.

Gordon

>but I received this error message:
>
>Error in "[<-"(`*tmp*`, , i, value = structure(list(), .Names = 
>character(0), row.names = c("1",  :
>         nothing to replace with
>
>I do apologies but I'm not yet confident in R so I can't understand how to 
>solve the problem...
>
>Maybe should I create another text file with two columns more for the 
>background and put zero in them?
>
>Does anyone know how to read this kind of data?
>
>Thanks in advance and sorry again if I made some trivial mistakes.
>Regards.
>
>Emanuele
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor