eBayes with 1 biological replicate

Question

[minni9234]

Does anyone know why the error occurred when I just make a comparison of 1 time point (2hHGF) sample and 1 time point (2hnoHGF) sample? The error is shown in yellow below. And how to fix the error?

> samples<-as.factor(c("2hnoHGF","2hHGF","8noHGF","8hHGF","24noHGF","24hHGF"))
> design<-model.matrix(~0+samples)
> colnames(design)<-levels(samples)
> fit<-lmFit(selExpr,design)
> contrast.matrix<-makeContrasts(2hHGF-2hnoHGF, levels=design)
> fits<-contrasts.fit(fit,contrast.matrix)

> ebfit<-eBayes(fits)

Error in ebayes(fit = fit, proportion = proportion, stdev.coef.lim = stdev.coef.lim,  : 
  No residual degrees of freedom in linear model fits

Answer 1

[dariober]

The conditions 2hHGF and 2hnoHGF have no replicates, so limma cannot estimate their dispersions. In other words, without replicates you can't tell whether an expression difference between 2hHGF and 2hnoHGF is due to the condition you want to test or it is due to the technical variability between arrays and samples.