Researchers have developed DeMixPert, a method for predicting how individual cells will respond to genetic perturbations. This is challenging because different cells may react differently to the same perturbation. DeMixPert decomposes the response into three components: the general reaction of a cell population, the specific effect of a particular perturbation, and random variation between cells. The method uses machine learning techniques to model these components and predict how individual cells will behave. According to the authors, DeMixPert outperforms other methods in predicting responses to unseen perturbations. --- Why it matters: This matters because accurately predicting how cells respond to genetic perturbations is crucial for understanding gene function and developing new treatments for diseases. DeMixPert's ability to model complex cellular behavior could lead to breakthroughs in fields like synthetic biology and regenerative medicine. Source: https://arxiv.org/abs/2608.23114