The authors present a mixed effects model of evolution (MEME) to detect sites under episodic diversifying selection in protein-coding genes. Traditional computational methods often fail to identify sites subject to episodic selection, which affects only a subset of lineages. MEME allows the distribution of the ω ratio to vary both within and between branches, providing a more flexible approach to modeling selective pressure. Using both empirical and simulated data, the authors demonstrate that MEME outperforms existing models in detecting episodic positive selection. They find that episodic selection is widespread and conclude that the number of sites experiencing positive selection may have been underestimated. The study highlights the importance of considering episodic selection in evolutionary biology and genetics, and provides a method to better identify such sites.The authors present a mixed effects model of evolution (MEME) to detect sites under episodic diversifying selection in protein-coding genes. Traditional computational methods often fail to identify sites subject to episodic selection, which affects only a subset of lineages. MEME allows the distribution of the ω ratio to vary both within and between branches, providing a more flexible approach to modeling selective pressure. Using both empirical and simulated data, the authors demonstrate that MEME outperforms existing models in detecting episodic positive selection. They find that episodic selection is widespread and conclude that the number of sites experiencing positive selection may have been underestimated. The study highlights the importance of considering episodic selection in evolutionary biology and genetics, and provides a method to better identify such sites.