Gene regulatory networks control functional phenotypes in cells and tissues. In complex tissues, these networks span cells and cell types to generate specialized functions, and are frequently dysregulated as both a cause and a consequence of disease. Technologies that leverage CRISPR–Cas9 to measure how gene silencing affects the expression level of other genes in the same cell, such as CROP-seq, Perturb-seq and CRISP-seq, are powerful tools to map intracellular genetic networks. High multiplexing can help to identify a network of regulatory effects. However, these methods require dissociating samples into single cells and are therefore limited to mapping intracellular gene networks. Also, they do not recover information about functional phenotypes regulated by each branch of the gene network.

Recording perturbation outcomes and gene expression in situ reveals the effects of the local environment on gene networks. For example, we showed the effects of local cell density and neighbourhood composition on the NF-κB pathway in macrophages and uncovered cell population-level regulation of this pathway. We also demonstrated the applicability of the approach to 3D systems in a tumour xenograft model.