[BioC] Limma package: problem with p-value and t

Julie Zhu julie.zhu at umassmed.edu
Thu Sep 17 22:35:26 CEST 2009


Dear Gordon,

I noticed that there is inconsistency between the t and p.value on the
topTable output using Limma package to analyze a dye swap two channel array
experiment. 
        logFC  AveExpr        t      P.Value  adj.P.Val        B
1401 3.216254 11.74137 22.49546 1.041470e-05 0.07874703 3.435097
1537 3.344735 11.88007 18.07448 2.702061e-05 0.07874703 2.951114

The two-tailed p-value for t=22.49546 with df=1 would be 0.028
2*pt(-abs(22.49546),df=1), but the P.Value in the above output is 1.04e-05.
Did I miss something? Could you please explain the inconsistency? Thanks!

Here is the design matrix and relevant code snippets.

   Dye mutant
   1     -1
   1     -1
   1      1
   1      1

design = cbind(Dye=1, mutant = c(-1,-1,1,1))
fit = lmFit(MA, design)
### MA is a MAlist object
fit3 = eBayes(fit)
topTable(fit3, number=2, coef="mutant", adjust="fdr")

Here is my R session information.

R version 2.9.0 (2009-04-17)
i386-apple-darwin8.11.1

locale:
en_US.UTF-8/en_US.UTF-8/C/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_2.18.0

Thank you very much for your help!

Best regards,

Julie



More information about the Bioconductor mailing list