A review on single-cell perturbation data analysis

Single-cell perturbation technologies combine external perturbations, such as gene editing or drug screening, with single-cell sequencing, providing detailed molecular profiles of cellular responses at single-cell re解. Computational analysis of these data aids in uncovering gene regulatory networks and drug mechanisms, as well as predicting the potential effects of unobserved gene or drug combinations. This review summarizes key single-cell perturbation technologies and highlights a growing range of methods for interpreting and predicting perturbation effects, extending from classical statistical models to machine learning and deep learning approaches. We also discuss the challenges and prospects, including perturbation atlases, multi-omics, spatial data, and causal learning.