Hi,

I have a bunch of genes wants to get gene-expression difference using Limma. In the output I would like to have Mean_Case + SD and Mean_Control + SD of each gene along with other parameters such as : gene-Name, log2FC, AveExpr, t, B, Pval, adj.P.Val.

any comments or suggestions!

Thanks,

Mahyar

You don't provide any code or experimental design (read the posting guide!) so I'll just make some data up:

g <- gl(5, 2)
y <- matrix(rnorm(20*length(g)), nrow=20)
design <- model.matrix(~0 + g)
fit <- lmFit(y, design)
fit <- eBayes(fit)

The group means are simply the values in fit$coefficients, due to the use of a no-intercept experimental design. The standard deviations can be obtained as fit$sigma, or sqrt(fit$s2.post) if you want the values after empirical Bayes shrinkage. Of course, if you're looking at the mean expression within each group, I suspect that you actually want the standard errors of the mean; these can be obtained as fit$stdev.unscaled * sqrt(fit$s2.post).

Everything else that you want is just the output of topTable() and can be obtained by calling the function with sort.by="none". Just cbind the results together to get a single data.frame object.