Fig. 3

Sets of DEGs and transcripts with significant RLRT for repeatedly simulated data. Sets of differentially expressed transcripts (DEGs) for both models and coincidences of transcripts with a significant variance component for imperfect blending. These were averaged over 100 repetitions of the simulated experiment based on the mouse design and variance components \({\sigma _{t}^{2}}=0.017,{\sigma _{b}^{2}}=0.094\) and \({\sigma _{z}^{2}}=2.7\) (all transcripts with effects for imperfect blending). The average counts of the sets of differentially expressed transcripts are labeled with ’DEG m1’ for the null model, ’DEG m2’ for the full model, ’VC significant’ for transcripts with a significant blending error variance, and ’DEG sim’ for the transcripts simulated as differentially expressed