[BioC] Limma time course analysis : determining positively and negatively reg genes

[Sean Davis]

On 10/19/05 12:43 PM, "Ken Termiso" <jerk_alert at hotmail.com> wrote:

> Hello,
> 
> I've analyzed a time course expt (on affy chips) with limma where there is
> one effect besides time. I've got 2 lists of significant diff expressed
> genes (one for each coefficient), have removed absent genes, and was
> wondering what the best way to get out which genes are positively affected
> and negatively effected...better yet, is there a way to calculate the actual
> fold changes for these genes for the two groups?

The M-value in the topTable output is the log2 fold change, so you can
filter based on that column to get at the genes that you like.

Sean