[R] qvalue

[Uwe Ligges]

alex lam (RI) wrote:
> Dear colleagues,
> 
> This is not strictly a R question, but I hope it is ok to ask on the
> list.
> I fed a vector of p-values from about 20 million anova tests to the
> package q-value and obtained this output:
> 
>> qsummary(asso_p.qvalue)
> 
> Call:
> qvalue(p = asso_p.vec)
> 
> pi0:    1
> 
> Cumulative number of significant calls:
> 
>         <1e-04 <0.001  <0.01 <0.025   <0.05    <0.1       <1
> p-value  22711  59067 262103 551316 1013241 1936797 20048103
> q-value   3554   5211   9275  12807   16951   24557 20048103
> 
> I can see that there are many hits with very small q-values ,say q
> <0.001, so to use that as cut off I have 5211 significant hits and can
> expect ~ 5 false postives. But my pi0 is 1. Does it mean that all my
> tests are true null and there are no significant results?

Well, yes, at least under H0 you would expect the above results in such 
a case.
Concluding without further information that *all* are non-significant is 
impossible, though. This requires some interpretation depending on where 
the tests some from...

Uwe Ligges



> Thanks for your help.
> Alex 
>  
> ------------------------------------
> Alex Lam
> PhD student
> Department of Genetics and Genomics
> Roslin Institute (Edinburgh)
> Roslin
> Midlothian EH25 9PS
> 
> Phone +44 131 5274471
> Web   http://www.roslin.ac.uk
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.