Multimodal single cell analyses reveal gene networks of planarian stem cell differentiation

Abstract Cell type identity is controlled by gene regulatory networks (GRNs), where transcription factors (TFs) regulate target genes (TGs) via open chromatin regions (OCRs), often specific to one or multiple cell types. Classic GRN discovery using perturbations is laborious and not easily scalable across the tree of life. Single-cell transcriptomics enables cell type-resolved gene expression analysis, but integrating perturbation data remains difficult. Here, we investigate planarian stem cell differentiation by integrating single-cell transcriptomics and chromatin accessibility data. The integrated analysis identifies gene networks matching known TF interactions and highlights TFs that may drive differentiation across multiple cell types. Our data reveals at least two major cell type supergroups linked by their regulatory logic, including alx3-1 + cells, comprising muscle, neurons and secretory cells, and hnf4 + cells, comprising gut phagocytes, goblet cells and parenchymal cells. We validated our data demonstrating high overlap between predicted targets and experimentally validated differentially regulated genes. Overall, our study integrates TFs, TGs and OCRs to reveal the regulatory logic of planarian stem cell differentiation, showcasing a comprehensive catalogue of GRN computational inferences that will be key to study this process.