[BioC] How to get adjusted p values after "global" fdr correction in decideTests (limma)

[Gordon K Smyth]

Dear Paolo,

What you say is correct.  The only way to get the adjusted p-values to go 
with method="global" is

   adj.p <- p.adjust(fit$p.value, method="BH")

where 'fit' is the fit object that you gave to decideTests().

Best wishes
Gordon

> Date: Wed, 23 Jul 2008 12:24:43 +0200
> From: Paolo Innocenti <paolo.innocenti at ebc.uu.se>
> Subject: [BioC] How to get adjusted p values after "global" fdr
> 	correction in decideTests (limma)
> To: AAA - Bioconductor <bioconductor at stat.math.ethz.ch>
>
> Hi all,
>
> I used the option "global" in decideTests(), in order to correct for FDR
>  taking into account my 3 contrasts simultaneously.
>
> The list of DE genes I got is longer than the one I get using "separate"
> option (and this should be ok, see here):
> http://thread.gmane.org/gmane.science.biology.informatics.conductor/17847
>
> Now, topTable and write.fit implement FDR adjustment with method
> "separate", so if I extract the p.values from these objects, some of
> them, in my situation, will be inevitably higher than 0.05.
>
> How can I get the correct p.values?
> Am I missing something obvious?
>
> Thanks in advance for any hints,
> paolo
>
>
>
> -- 
> Paolo Innocenti
> Department of Animal Ecology, EBC
> Uppsala University
> Norbyv?gen 18D
> 75236 Uppsala, Sweden