Simulated multi-condition scRNA-seq data

This dataset contains simulated single-cell RNA sequencing data generated using scDesign3 to benchmark scCausalVI and other computational methods for analyzing perturbational single-cell experiments. The simulation is based on an IFN-beta stimulated PBMC dataset and includes 6,852 cells across four cell types (B cells, CD4+ naive T cells, CD14+ monocytes, and activated T cells) and three experimental conditions. The control group maintains natural cellular variations without perturbation effects, while two treatment groups introduce cell-type-specific condition effects: a 1.5-fold effect exclusively on CD4+ naive T cells in treatment 1, and a 2-fold effect exclusively on B cells in treatment 2. Each cell type-condition combination contains 571 cells, with expression profiles for 1,000 highly variable genes. This controlled simulation design enables systematic evaluation of methods that aim to disentangle inherent cellular heterogeneity from treatment-induced changes. The data is provided in standard single-cell analysis formats with accompanying metadata specifying cell type labels and experimental condition assignments.