[BioC] topTable in limma

Gordon Smyth smyth at wehi.EDU.AU
Thu Sep 20 03:53:30 CEST 2007


>Date: Tue, 18 Sep 2007 17:40:26 -0400 (EDT)
>From: Brooke LaFlamme <bal44 at cornell.edu>
>Subject: [BioC] topTable in limma
>To: bioconductor at stat.math.ethz.ch
>
>Sorry, I had one more question!
>
>If I have more that one contrast in my fit (after running 
>ebayes(fit)), and I use this to produce a table of top genes using 
>topTable(), does the list of genes produced simply correspond to the 
>genes with the highest probability of being significant in one or 
>more contrasts, up to N genes?

If you specify a particular contrast, topTable will rank by p-values 
and B-statistics for that contrast. If you don't, it will rank on 
F-statistics, which is a test of overall significance for all (one or 
more of) the contrasts.

Best wishes
Gordon

>This is what I assumed to be the case, but I'm not certain.
>
>Thanks again!
>
>Brooke L.



More information about the Bioconductor mailing list