Better handling of the interbatch replicates...

Created by: ireneg

... which does not seem to make much of a difference at all.

Used blocking and duplicateCorrelation on the basis of past experience with it and looking at biostars and bioconductor posts. The closest questions I could find was this one but note that blocking isn't possible for them because it's already being used for the individual pairing, so not an identical situation. Ultimately it did not make much of a difference to the baseline approach, so it is safe to continue no matter what.

Sanity checks courtesy of Gordon in John Blischak's post from ages ago.

Model matrix contains only granulocyte estimates just for ease of running, but needs to be updated.