I'm working with some DNA methylation data and would like to use either Jaffe et al.'s bumphunter algorithm as implemented in the minfi package or the DMRcate package to identify differentially methylated regions. My study is a bit more complex as I have repeated measures (two time points) and therefore am running into some trouble specifying a design matrix. I would like to use a linear mixed effect model (lmer) to account for the repeated measures. Does anyone have experience with this? Any guidance would be much appreciated!
I would like to specify my design matrix to look something like this...