[BioC] limma reading Agilent One-Color Data

elliott harrison e.harrison at epistem.co.uk
Thu Nov 1 18:23:01 CET 2007

Hi BioC,

In a discussion with the same title as this Gordon advised the following
working with one colour data.

1. Use read.maimages() with dummy arguments for R & Rib
2. Background correct as usual using background Correct()
3. Normalize the GR$G matrix using normalizeBetweenArrays()
4. Use log2(RG$G) as input to lmFit() etc.

In the 4th step log2(RG$G) as input if I use normalizeBetweenArrays
first there is no $G component of the object to use as the input of

That issue could be completely unrelated to the next.

I'm trying to find differentially expressed genes between 2 arrays

> R25 = cbind(R[,2], R[,5])
Have tried both of
> RR25 <- normalizeBetweenArrays(R25$R, method="quantile")
> RR25 <- normalizeBetweenArrays(R25, method="quantile")

Then continue through to
> g <- paste(R25$targets$SlideNumber)
> g <- factor(g)
> design <- model.matrix(~0+g)
> colnames(design) <- levels(g)
> fit <- lmFit(log2(R25), design)
>  cont.matrix<- makeContrasts(OneVFour="A9802-A9811",levels=design)
> fit2  <- contrasts.fit(fit, cont.matrix)

> fit2  <- eBayes(fit2)
Error in ebayes(fit = fit, proportion = proportion, stdev.coef.lim =
stdev.coef.lim) : 
        No residual degrees of freedom in linear model fits

Whatever I pass into normalizeBetweenArrays be it R25$R or R25 I have
nothing to analyse by the time I get to the eBayes function.

Is this related to me not following what Gordon advised correctly?
I am doing something wrong in the differential expression analysis?
Is it duff data?

Any ideas?



This message has been scanned for viruses by BlackSpider...{{dropped:3}}

More information about the Bioconductor mailing list