[BioC] A limma question

[Gordon Smyth]

>[BioC] A limma question.
>Johan Lindberg johlin at kth.se
>Fri Apr 21 16:56:06 CEST 2006
>
>I think this is kind of a stupid question but I cant get it right so 
>I will risk my hide.
>
># First I do a topTable as usual, but without adjustment for multiple testing.
>topt.dc <- 
>topTable(fit=ebobj,number=length(ebobj$coef),adjust.method="none",sort.by="t")
>
> > topt.dc[1:3,]
>               M          A         t      P.Value         B
>20861 -1.025634 -1.0256342 -14.87215 1.285609e-10 13.716847
>680   -1.889194 -1.8891945 -14.58354 1.711049e-10 13.493043
>24876 -1.861192 -1.8611920 -14.30543 2.264364e-10 13.271669
>
># The degrees of freedom for the moderated t-test is if I am not mistaken
>df.dc <- ebobj$df.prior + ebobj$df.residual
>
># If you look at the toptable the P-value for the first gene is 1.285609e-10
>and the t-value is -14.87215.
>
># but if I try to get the t-value from the p-value by
>qt(p=topt.dc[1,4], df=df.dc[idx.vec[1]],lower.tail = TRUE)
>-14.18081
>
>#which differs from -14.87215. The difference is bigger for the second gene.
># -13.9039 vs -14.58354
>
>What am I doing wrong?

You've forgotten that the p-values are two-sided.

Best wishes
Gordon

>Best regards
>// Johan Lindberg
>
>****************************************************************************
>Johan Lindberg
>Royal Institute of Technology
>AlbaNova University Center
>Stockholm Center for Physics, Astronomy and Biotechnology
>School of Molecular Biotechnology
>Department of Gene Technology
>Visiting address:
>Roslagstullsbacken 21, Floor 3
>106 91 Stockholm, Sweden
>Delivering address:
>Roslagstullsbacken 30B
>104 06 Stockholm, Sweden
>Phone (office) +46 8 553 783 44
>Fax + 46 8 553 784 81
>  <http://www.ktharray.se/> http://www.ktharray.se/
>  <http://www.arrayadvice.se/> http://www.arrayadvice.se/
>****************************************************************************