RADMeth DMR in IGV:
Looks like I may have spoken too soon. I loaded up the RADMeth produced DMRs in IGV and the DMRs didn’t even remotely line up to sites with methylation in the Day 10 samples. This may be an artifact of the merging process, or something being wonky with the model. Off the top of my head, potential oddities for the model may include a lack of nearby neighbors for the p-value adjusting step, or some oddity in trying to calculate DMRs from a very sparse set of loci.
I’m going to try to figure out how to load up either the regression or the adjustment steps in IGV and see if that shines any light on the problem.