[BioC] Limma and negative p-values

Naomi Altman naomi at stat.psu.edu
Thu Mar 15 16:37:49 CET 2012


I have posted this several times, but it has been a while so I will 
do it again.

Suppose p=.001 and you have 10000 genes.  Then you expect 10 
significant genes just by chance.  So if you observe 9 or 10 
significant genes at this level, your FDR should be 100%.  If you 
observe 11, your FDR should be about 10/11=0.91.  Etc.

This does not mean that you don't have any significant genes.  It 
means your experiment did not have enough power to detect them.

--Naomi


At 10:31 AM 3/15/2012, Ovokeraye Achinike-Oduaran wrote:
>Hi,
>
>Excuse the language but I say shady because they all range from 
>~0.918 - 0.999.
>
>On Thu, Mar 15, 2012 at 3:54 PM, Steve Lianoglou
><mailinglist.honeypot at gmail.com> wrote:
> > Hi,
> >
> > On Thu, Mar 15, 2012 at 9:50 AM, Ovokeraye Achinike-Oduaran
> > <ovokeraye at gmail.com> wrote:
> >> Hi all,
> >>
> >> Sorry...a recall of my previous mail...the p-values are fine(or not)
> >> but not negative...the corrected p-values, however, look very shady.
> >> Any suggestions there?
> >
> > Shady?
> >
> > Need to elaborate, please.
> >
> > -steve
> > --
> > Steve Lianoglou
> > Graduate Student: Computational Systems Biology
> >  | Memorial Sloan-Kettering Cancer Center
> >  | Weill Medical College of Cornell University
> > Contact Info: http://cbio.mskcc.org/~lianos/contact
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at r-project.org
>https://stat.ethz.ch/mailman/listinfo/bioconductor
>Search the archives: 
>http://news.gmane.org/gmane.science.biology.informatics.conductor



More information about the Bioconductor mailing list