Hi, I have two groups of affymetrix microarray data, one group contains more than 10 normal microarrays, while another group only have 1 cancer microarray. I wonder whether limma could be used to identify differential-expressed genes by comparing these two groups? I am concerned about the accuracy of limma since the variance of the cancer data could not be estimated due to the sample limitation. Thanks, Lei -- output of sessionInfo(): NA -- Sent via the guest posting facility at bioconductor.org.