My RNA-Seq experiment is comprised of two technical replicates each of a normal and affected phenotype across four time points (yielding 16 samples). My understanding of the EBSeq-HMM algorithm is that genes are considered differentially expressed only with respect to time points. Is there a way to accommodate treatment information? Is there a conceptual mistake in running the algorithm for each treatment separately and retrieving genes with expression paths exclusive to the affected phenotype?