[BioC] limma: fold-changes corrected for dye-effect

[Naomi Altman]

When the design is dye-balanced, the dye effect does not affect the 
estimated contrasts, but it will affect the estimate of variance and 
hence the p-values.

--Naomi

At 10:11 AM 4/22/2007, karsten_hokamp at sfu.ca wrote:
>Hi list,
>
>In the limma userguide it says at the end of section "8.1.2 Dye Swaps"
>(page 34): "The fold changes and significant tests in this list are
>corrected for dye-effects."
>
>When I process microarray data with and without dye-effect specified I can
>see the p-values being corrected, but the logFC values as reported by
>topTable stay the same.
>
>Below are the steps I carried out on a small artificial sample data set:
>
> > files = c('val1', 'val2', 'val3', 'val4')
> > RG = read.maimages(files, columns=list(G="F1",R="F2",Gb="B1",Rb="B2"),
>+ annotation=c("Spot"))
>Read val1
>Read val2
>Read val3
>Read val4
> > ma = MA.RG(RG, bc.method="none")
> > ma$M
>      val1 val2 val3 val4
>[1,]   -3    1   -1    1
>[2,]    1   -2    1   -1
>[3,]    4   -4    2   -4
>[4,]    0    0    0    1
>[5,]    4   -6    4   -6
> >
> > design <- cbind(DyeEffect=1,MUvsWT=c(1,-1,1,-1))
> > design2 <- cbind(MUvsWT=c(1,-1,1,-1))
> > fit <- lmFit(ma, design)
> > fit <- eBayes(fit)
> > fit2 <- lmFit(ma, design2)
> > fit2 <- eBayes(fit2)
> > topTable(fit, coef="DyeEffect")
>   logFC AveExpr           t      P.Value    adj.P.Val         B
>5 -1.00   5.500 -124.500939 1.338072e-05 0.0000669036  4.801725
>1 -0.50   1.750   -1.091256 3.731143e-01 0.3732967729 -8.968102
>3 -0.50   4.250   -1.091256 3.731143e-01 0.3732967729 -8.968102
>2 -0.25   7.375   -1.090753 3.732968e-01 0.3732967729 -8.968621
>4  0.25   7.125    1.090753 3.732968e-01 0.3732967729 -8.968621
> > topTable(fit, coef="MUvsWT")
>   logFC AveExpr          t      P.Value    adj.P.Val         B
>5  5.00   5.500 622.504696 2.892709e-07 1.446354e-06  7.651719
>3  3.50   4.250   7.638791 9.962634e-03 2.490659e-02 -4.415323
>2  1.25   7.375   5.453766 2.148516e-02 3.580860e-02 -5.491825
>1 -1.50   1.750  -3.273768 6.461363e-02 8.076704e-02 -7.008591
>4 -0.25   7.125  -1.090753 3.732968e-01 3.732968e-01 -9.206043
> > topTable(fit2)
>   logFC AveExpr          t      P.Value    adj.P.Val         B
>5  5.00   5.500 10.6609570 1.549942e-26 7.749712e-26 49.311955
>3  3.50   4.250  7.4626699 8.478657e-14 2.119664e-13 20.722709
>1 -1.50   1.750 -3.1982871 1.382466e-03 2.304109e-03 -1.700230
>2  1.25   7.375  2.6652393 7.693357e-03 9.616696e-03 -3.241807
>4 -0.25   7.125 -0.5330479 5.940005e-01 5.940005e-01 -6.605248
> > sessionInfo()
>R version 2.4.1 (2006-12-18)
>i686-pc-linux-gnu
>
>locale:
>LC_CTYPE=en_US.UTF-8;LC_NUMERIC=C;LC_TIME=en_US.UTF-8;LC_COLLATE=en_US.UTF-8;LC_MONETARY=en_US.UTF-8;LC_MESSAGES=en_US.UTF-8;LC_PAPER=en_US.UTF-8;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US.UTF-8;LC_IDENTIFICATION=C
>
>attached base packages:
>[1] "stats"     "graphics"  "grDevices" "utils"     "datasets"  "methods"
>[7] "base"
>
>other attached packages:
>    limma
>"2.9.17"
>
>
>Am I doing something wrong?
>
>Thanks for any help,
>
>         Karsten
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at stat.math.ethz.ch
>https://stat.ethz.ch/mailman/listinfo/bioconductor
>Search the archives: 
>http://news.gmane.org/gmane.science.biology.informatics.conductor

Naomi S. Altman                                814-865-3791 (voice)
Associate Professor
Dept. of Statistics                              814-863-7114 (fax)
Penn State University                         814-865-1348 (Statistics)
University Park, PA 16802-2111