as you have a problem of class comparison between two classes I suggest you to read the user guide of the limma package of Bioconductor, which is a popular package for identification of differentially expressed genes (http://www.bioconductor.org/packages/release/bioc/vignettes/limma/inst/doc/usersguide.pdf). You can focus on section 9.2 if you are working with one-color microarrays.
By the way, you have to create a two-level factor to perform the comparison:
# build the design matrix
library(limma)
yourfactor <- c(rep("RAL", 30),rep("ZI", 5))
design <- model.matrix(~ 0 + yourfactor)
colnames(design) <- gsub("yourfactor", "", colnames(design)) # to simplify the colnames of design
# perform the comparison
fit <- lmFit(data, design) # data is your gene expression matrix
contrast.matrix <- makeContrasts(RAL-ZI, levels=design)
fit2 <- contrasts.fit(fit, contrast.matrix)
fit2 <- eBayes(fit2)
# summarize the results of the linear model
results <- topTable(fit2, number=nrow(data), adjust.method="BH")
Be careful that samples in the expression matrix and sample labels in your factor are in the same order. To avoid this kind of problem I suggest you to create an ExpressionSet object (http://www.bioconductor.org/packages/release/bioc/html/Biobase.html) which is very useful for manipulating gene expression data.
I hope this was helpful,
Best.
Matteo