I have a data set with both paired samples and confounding factors. I want to use limma to find the differential expressed genes, but I don't know how can I design the design matrix. Could anyone give some suggestions? The data looks like this,
The variable Condition, namely Treated or Control, is the main effect need to be explored. Pair_Name reflects the pairing of the samples, namely the samples in pair 1 from the same donator, samples in pair 2 from the 2nd donator, pair 3 from the 3rd. Soure is the source of the sample, some from the tissue of interest, while some from blood, which is a discrete variable, and also a confounding factor. The last variable Days is a continuous variable and another confounding factor.
What I want to do is to select the significantly differential expressed genes between the 2 conditions Treated and Control, while considering the effect of sample pairing and the confounding factors Source and Days. It seems very complicated now and I don't know how to design the design matrix and contrast matrix properly. Could anyone give me some suggestions? Thank you so much!