LIMMA error: No residual degrees of freedom

0

Entering edit mode

SAURIN ★ 1.1k

@saurin-799

Last seen 10.2 years ago

Hi , I have 4 CEL files and each one is in different Group. I am doing ReadAffy() and RMA > pData(myRMA); sample pheno1 pheno2 A 1 Group1 1 B 2 Group2 2 C 3 Group3 3 D 4 Group4 4 > design <- model.matrix(~ 0 +factor(c(1,2,3,4))); > colnames(design) <- c("Group1","Group2","Group3","Group4"); > fit <- lmFit(myRMA,design); > contrast.matrix1 <- makeContrasts(Group2-Group1,Group3-Group4,Group2-Group3,Group4-Group1, Group3-Group1,Group4-Group2,levels = design); > fitM1 <- contrasts.fit(fit,contrast.matrix1); > fitM1<- eBayes(fitM1); Error in ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim) : No residual degrees of freedom in linear model fits Any suggestion will be great, Thank you, Saurin __________________________________ The all-new My Yahoo! - What will yours do?

• 6.2k views

ADD COMMENT • link updated 20.0 years ago by James W. MacDonald 67k • written 20.0 years ago by SAURIN ★ 1.1k

1

Entering edit mode

James W. MacDonald 67k

@james-w-macdonald-5106

Last seen 2 days ago

United States

Saurin Jani wrote: > Hi , > > I have 4 CEL files and each one is in different Group. > I am doing ReadAffy() and RMA With only one sample per group you cannot fit a model because you have no replication (hence, no residual degrees of freedom). You will only be able to compute fold changes. > > >>pData(myRMA); > > sample pheno1 pheno2 > A 1 Group1 1 > B 2 Group2 2 > C 3 Group3 3 > D 4 Group4 4 > > > >>design <- model.matrix(~ 0 +factor(c(1,2,3,4))); >>colnames(design) <- > > c("Group1","Group2","Group3","Group4"); > >>fit <- lmFit(myRMA,design); >>contrast.matrix1 <- > > makeContrasts(Group2-Group1,Group3-Group4,Group2-Group3,Group4-Group 1,Group3-Group1,Group4-Group2,levels > = design); > >>fitM1 <- contrasts.fit(fit,contrast.matrix1); >>fitM1<- eBayes(fitM1); > > Error in ebayes(fit = fit, proportion = proportion, > stdev.coef.lim = stdev.coef.lim) : > No residual degrees of freedom in linear model > fits > > > Any suggestion will be great, > > Thank you, > Saurin > > > > > > __________________________________ > > The all-new My Yahoo! - What will yours do? > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor -- James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109

ADD COMMENT • link 20.0 years ago James W. MacDonald 67k

0

Entering edit mode

Hi, If I use more than 4 files for 4 groups..(right now I am using 16 CEL files, which generated on diff. days on set of 4), I get zero in Group1-Group2 or any contrast matrix venn diagram. contrast.matrix <- makeContrasts(group4-group1,group3-group1,group2-group1,group4-group2, group3-group2,group4-group3,levels = design); fit2 <- contrasts.fit(fit,contrast.matrix); fit2 <- eBayes(fit2); results <- decideTests(fit2); vennDiagram(results); I get zero in any comparision,Which means my data is highly variable or something wrong in my design matrix? Thank you, Saurin --- "James W. MacDonald" <jmacdon@med.umich.edu> wrote: > Saurin Jani wrote: > > Hi , > > > > I have 4 CEL files and each one is in different > Group. > > I am doing ReadAffy() and RMA > > With only one sample per group you cannot fit a > model because you have > no replication (hence, no residual degrees of > freedom). You will only be > able to compute fold changes. > > > > > > > >>pData(myRMA); > > > > sample pheno1 pheno2 > > A 1 Group1 1 > > B 2 Group2 2 > > C 3 Group3 3 > > D 4 Group4 4 > > > > > > > >>design <- model.matrix(~ 0 +factor(c(1,2,3,4))); > >>colnames(design) <- > > > > c("Group1","Group2","Group3","Group4"); > > > >>fit <- lmFit(myRMA,design); > >>contrast.matrix1 <- > > > > > makeContrasts(Group2-Group1,Group3-Group4,Group2-Group3,Group4-Group1, Group3-Group1,Group4-Group2,levels > > = design); > > > >>fitM1 <- contrasts.fit(fit,contrast.matrix1); > >>fitM1<- eBayes(fitM1); > > > > Error in ebayes(fit = fit, proportion = > proportion, > > stdev.coef.lim = stdev.coef.lim) : > > No residual degrees of freedom in linear > model > > fits > > > > > > Any suggestion will be great, > > > > Thank you, > > Saurin > > > > > > > > > > > > __________________________________ > > > > The all-new My Yahoo! - What will yours do? > > > > _______________________________________________ > > Bioconductor mailing list > > Bioconductor@stat.math.ethz.ch > > https://stat.ethz.ch/mailman/listinfo/bioconductor > > > -- > James W. MacDonald > Affymetrix and cDNA Microarray Core > University of Michigan Cancer Center > 1500 E. Medical Center Drive > 7410 CCGC > Ann Arbor MI 48109 > __________________________________ The all-new My Yahoo! - What will yours do?

ADD REPLY • link 20.0 years ago SAURIN ★ 1.1k

0

Entering edit mode

Saurin Jani wrote: > Hi, > > If I use more than 4 files for 4 groups..(right now I > am using 16 CEL files, which generated on diff. days > on set of 4), I get zero in Group1-Group2 or any > contrast matrix venn diagram. > > > contrast.matrix <- > makeContrasts(group4-group1,group3-group1,group2-group1,group4-group 2,group3-group2,group4-group3,levels > = design); > fit2 <- contrasts.fit(fit,contrast.matrix); > fit2 <- eBayes(fit2); > > results <- decideTests(fit2); > > vennDiagram(results); > > > I get zero in any comparision,Which means my data is > highly variable or something wrong in my design > matrix? Without seeing the design matrix it is hard to say. However, if everything is set up correctly and you don't get any genes, that would indicate that none of your fdr-adjusted p-values are less than 0.05. This may indicate that your data are highly variable. You say that the chips were generated on different days. How much time elapsed between each chip run? Are we talking weeks or months here? Also, did you switch from the Enzo IVT kit to the Affy kit? If so, that is a problem too. > > Thank you, > Saurin > -- James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109

ADD REPLY • link 20.0 years ago James W. MacDonald 67k

0

Entering edit mode

Hi, Thank you for your reply. My data is generated 24 hrs of time e.g. (day1) 4 files, (day2) 4 files like that. > design <- model.matrix(~ 0 +factor(c(1,1,1,1,2,2,2,2,3,3,3,3,4,4,4,4)); > colnames(design) <- levels(myRMA$pheno1); > design Group1 Group2 Group3 Group4 1 1 0 0 0 2 1 0 0 0 3 1 0 0 0 4 1 0 0 0 5 0 1 0 0 6 0 1 0 0 7 0 1 0 0 8 0 1 0 0 9 0 0 1 0 10 0 0 1 0 11 0 0 1 0 12 0 0 1 0 13 0 0 0 1 14 0 0 0 1 15 0 0 0 1 16 0 0 0 1 attr(,"assign") [1] 1 1 1 1 attr(,"contrasts") attr(,"contrasts")$"factor(c(1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 4))" [1] "contr.treatment" I am using Affy kit for over 1 year. I think as well, my data is highly variable even which has only 24 hrs time between them. Thank you, Saurin --- "James W. MacDonald" <jmacdon@med.umich.edu> wrote: > Saurin Jani wrote: > > Hi, > > > > If I use more than 4 files for 4 groups..(right > now I > > am using 16 CEL files, which generated on diff. > days > > on set of 4), I get zero in Group1-Group2 or any > > contrast matrix venn diagram. > > > > > > contrast.matrix <- > > > makeContrasts(group4-group1,group3-group1,group2-group1,group4-group2, group3-group2,group4-group3,levels > > = design); > > fit2 <- contrasts.fit(fit,contrast.matrix); > > fit2 <- eBayes(fit2); > > > > results <- decideTests(fit2); > > > > vennDiagram(results); > > > > > > I get zero in any comparision,Which means my data > is > > highly variable or something wrong in my design > > matrix? > > Without seeing the design matrix it is hard to say. > However, if > everything is set up correctly and you don't get any > genes, that would > indicate that none of your fdr-adjusted p-values are > less than 0.05. > This may indicate that your data are highly > variable. > > You say that the chips were generated on different > days. How much time > elapsed between each chip run? Are we talking weeks > or months here? > Also, did you switch from the Enzo IVT kit to the > Affy kit? If so, that > is a problem too. > > > > > > Thank you, > > Saurin > > > > -- > James W. MacDonald > Affymetrix and cDNA Microarray Core > University of Michigan Cancer Center > 1500 E. Medical Center Drive > 7410 CCGC > Ann Arbor MI 48109 >

ADD REPLY • link 20.0 years ago SAURIN ★ 1.1k

Login before adding your answer.