[BioC] Q-values way smaller than P-values! Is that kosher?

[Joshua Banta [guest]]

Dear listserv,

Consider the following vector of P-values:

p.values <- c(
0.266845129, 
0.353675275,
0.030585143,
0.791285527,
0.60805832,
0.433466716,
0.039369439,
0.48918122,
0.740626586
)

When I run the qvalue function of the qvalue package, I get a rather surprising outcome: the q-values are much smaller than the corresponding p-values!

See the following code below:

source("http://bioconductor.org/biocLite.R")
biocLite("qvalue")
library(qvalue)

q.values <- qvalue(p.values)$qvalue

comparison <- cbind(p.values, q.values)

colnames(comparison) <- c("p.values", "q.values")

comparison

        p.values    q.values
 [1,] 0.26684513 0.037111560
 [2,] 0.35367528 0.037111560
 [3,] 0.03058514 0.008960246
 [4,] 0.79128553 0.040020397
 [5,] 0.60805832 0.039540110
 [6,] 0.43346672 0.037111560
 [7,] 0.03936944 0.008960246
 [8,] 0.48918122 0.037111560
 [9,] 0.74062659 0.040020397

So what is happening here? Can I trust the q-values? In each case they are substantially larger than the P-values. Especially troubling to me are rows 4, 5, and 9.

Thanks in advance for any information/advice!
-----------------------------------
Josh Banta, Ph.D
Assistant Professor
Department of Biology
The University of Texas at Tyler
Tyler, TX 75799
Tel: (903) 565-5655
http://plantevolutionaryecology.org

 -- output of sessionInfo(): 

R version 3.0.1 (2013-05-16)
Platform: x86_64-apple-darwin10.8.0 (64-bit)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] limma_3.16.8         BiocInstaller_1.10.4 qvalue_1.34.0       
[4] gsubfn_0.6-5         proto_0.3-10         emma_1.1.2          

loaded via a namespace (and not attached):
[1] tcltk_3.0.1 tools_3.0.1

--
Sent via the guest posting facility at bioconductor.org.