Hi,

I have lipidomic data of about 500 different lipid species, but only 4 samples in each group. The concentration of different lipids within each sample varies a lot. Can I use the eBayes() function to get the moderated standard error and make volcano plot? Or what is the best way to adjusted the standard error and conduct a statistical test?

Theoretically, you might log transformation the concentrations, then quantile normalize, and then analyse like microarray data in limma. The limma package is often used to analyse data other than gene expression values. However I have not seen or tried to analyse lipid concentration data, so I have no practical experience.