[BioC] What does this mean in Limma?

[Gordon K Smyth]

Dear Davide,

The short answer is that you shouldn't be interpreting the fit3 results 
because they are probably not want you want.

You have fitted a linear model in which there is no intercept, but instead 
three coefficients representing the means of the three groups.  If you 
simply display the results of topTable(fit3), then limma will test whether 
any of the coefficients are equal to zero.  This will select genes that 
have mean log-intensities in any group greater than zero.  So you are 
simply getting a list of highly expressed genes, not genes that are 
differently expressed between your groups.

Best wishes
Gordon


> Date: Fri, 1 Feb 2013 22:12:35 +0100
> From: Davide Cittaro <cittaro.davide at hsr.it>
> To: "bioconductor at r-project.org" <bioconductor at r-project.org>
> Subject: [BioC] What does this mean in Limma?
>
> Hi all,

> I'm performing an analyis in limma with 3 groups. Normally I would set 
> up a simple analysis with 3 contrasts:
>
> groups <- factor(c(A,A,A,B,B,B,C,C,C))
> design <- model.matrix(~0+group)
> contrasts <- makeContrasts(B-A, C-A, C-B, levels=design)
> fit <- lmFit(y, design)
> fit2<-eBayes(contrasts.fit(fit, contrasts))
>
> Because of an error, I've didn't set the matrix and performed
>
> fit3<-eBayes(fit)
>
> In which way I should interpret the results in fit3 (which, by the way, 
> are really interesting)?
>
> Thanks
>
> d
>

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}