If you've ever felt like you are in over your head trying to define #RStats linear models for your -omics experiments, this paper does an excellent job of showing the reasoning why the model is defined that way, what contrasts you want to define, and what is actually being tested.
A guide to creating design matrices for gene expression experiments
https://f1000research.com/articles/9-1444/v1
These apply to more than just gene expression #transcriptomics. Work for #proteomics and #metabolomcs too.