limma applied to unbalanced micro-array experiment

Question

[ewre]

Hello,

I have a microarray gene expression dataset which consists of 5 treated samples and only 1 control samples. I know that this kind of unbalanced design is not good for differential gene expression analysis. what I want to know is that which kind of differential gene expression method is most proper for this dataset?
1. a simple t test with the hypothesis that the data has a normal distribution
2. non parametric tests like rank test or rankprod
3. more advanced methods like limma

thank you for your time.

Answer 1

[Devon Ryan]

The most appropriate method is limma (regardless of whether you have a balanced design or not).

[ewre]

Hi Devon Ryan. thank you for your reply. Since I have only one sample as control, will limma handle this kind of situiation properly?

[Devon Ryan]

Yes it will.