[BioC] rownames in topTable and posterior probability

[Gordon K Smyth]

> Date: Fri, 25 Feb 2005 11:33:57 +0000
> From: Adaikalavan Ramasamy <ramasamy at cancer.org.uk>
> Subject: [BioC] rownames in topTable and posterior probability
> To: BioConductor mailing list <bioconductor at stat.math.ethz.ch>
>
> topTable() in limma packages returns the row numbers and it would be
> useful to see the row names (i.e. Affy ID) instead.

topTable() already does this, as can be seen from the case studies in the User's Guide.  For
example, the Ecoli and Estrogen case studies show top tables with Affy IDs included automatically.

> I am aware that I
> may supply the 'genelist' argument but help("topTable") says :
>
> 	genelist: data frame or character vector containing gene
> 	 	  information. If not specified, this will be taken
> 		  from the 'genes' component of 'fit'.

This is not normally required.  Have you not looked at the examples in the User's Guide?

> My usual usage is as follows
>
> fit  <- lmFit(onebatch, design)
> fit2 <- contrasts.fit( fit, cont.m )
> fit2 <- eBayes(fit2)
>
> topTable(fit2)
>                M         t      P.Value         B
> 12358 -0.8636721 -7.585040 3.808592e-05 11.958486
> 48048 -0.8765769 -7.400093 7.417187e-05 11.363358
> ...
>
> But I found no genes component in fit via fit$genes or str(fit). Same
> goes for fit2. I doing something wrong ?

You haven't told use what sort of object 'onebatch' is.  If onebatch is an object from which gene
names can be extracted, e.g., an exprSet, then they should be included in 'fit'.  If 'onebatch'
doesn't include gene names, you can't really expect them to be in fit.

Gordon

> Thank you.
>
> Regards, Adai