[BioC] Very low P-values in limma

[Thomas Hampton]

I too find it very odd that a multiple hypothesis correction  
apparently reduces P values.  One possibility is that
the duplicate spot weighting in limma is somehow doing a really  
superior job compared to a simple t test
that might view duplicates on different arrays as the same thing as  
duplicates on the same array.

Tom
On Oct 21, 2009, at 1:36 PM, Paul Geeleher wrote:

> Hi folks, I'm analyzing microRNA data using limma and I'm wondering  
> about
> the validity of the p-values I'm getting out. Its a simple 'Group A  
> Vs Group
> B' experimental design. 4 arrays in one group, 3 in the other and 4
> duplicate spots for each miRNA on each array.
>
> The lowest adjusted p-values in the differential expression analysis  
> are in
> the region of 10^-7.
>
> Its been pointed out to me that plugging the point values from each  
> sample
> into a regular t-test you get p=0.008, which then also needs to take  
> the
> multiple test hit. Can anybody explain why limma is giving me such  
> lower
> values and if they are valid?
>
> I can provide more information if required.
>
> Thanks,
>
> Paul.
>
> -- 
> Paul Geeleher
> School of Mathematics, Statistics and Applied Mathematics
> National University of Ireland
> Galway
> Ireland
>
> 	[[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor