i started preprossesing a small affymetrix microarray dataset. Due to some technical problems, i aquired only one control sample and 4 different samples representing 4 different biological susbstances as treatments(5 CEL files total). Except from quality control, could i use some method for DE expression, because due to the number and the phenotype of the data, i think it is not possible, even with limma. So my question is should i wait for gaining some more samples or there is some alternative? i have heard of rank products algorithm but i have never used it.