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

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_U S.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