Linear models enable powerful differential activity analysis in massively parallel reporter assays.

Together, these results inform recommendations for differential analysis, general group comparisons, and power analysis and will help improve design and analysis of MPRA experiments.