[BioC] p-values

[Renate van Dorp]

Hi there,

I am analyzing microarray data. We use the 2-step mixed model analyses 
proposed by Wolfinger et al (2001). One phenomena that we observe is that 
after analyses we plot the p-values. The p-values do not follow an uniform 
distribution with more differentially expressed genes at the 5 % 
significance level. We observe fewer genes at the 5% significance level and 
the percent of observations at higher p-values increase. I am curious to 
know if other groups have seen this or other types of distributions for the 
p-value. How do you deal with this and how do you analyze the data?

Thanks,

Renate

Renate T.E. van Dorp, PhD.
Center for Animal Functional Genomics
Department of Animal Science
B212 Anthony Hall
Michigan State University
East Lansing, MI 48824
Office (517) 432-1379
Lab (517) 353-9815
FAX (517) 353-1699