Extended Data Fig. 7: Transcriptional responses to genetic perturbations across targets and cell types.

a, A heatmap displaying the number of differentially expressed genes (DEGs) (q < 0.05) for each broad cell type, across all perturbations. Numbers are displayed in log₁₀(x + 1). b, A scatter plot comparing the mean number of DEGs (q < 0.05) for each cell type across perturbations to the mean number of cells per embryo (pearson R = 0.62). c, A heatmap displaying the normalized estimates from DEG testing in periderm cells across all perturbations (q < 0.05, n = 3206 genes). “Gene”-mut refers to null mutants (or -/-) rather than crispants. d, UMAP plots in which all neural progenitor cells are grey, and blue cells are control cells that are determined with the Getis–Ord test to have neighbors depleted for the perturbed cell type, termed “cold spots” for a selected set of perturbations known to affect hindbrain development. e, A heatmap displaying the normalized enrichment scores (NES) from a Gene Set Enrichment Analysis (GSEA)^72,77 with the hallmark gene set on averaged, ranked estimates from differential expression testing across cell types for each perturbation. Only pathways with at least one significant enrichment are displayed (p-adj < 0.05, number of random gene sets with the same or larger value divided by the total number of generated sets, followed by multiple testing correction), and the color corresponds to the magnitude and direction of each significant enrichment; non-significant are white. Perturbations are annotated by whether they are null mutants (mutant) or F0 CRISPR/Cas9-injected (crispant). f, Scatter plots displaying the number of significant, differentially expressed genes between perturbed cells and control cells (y-axis), versus the absolute fold change in cell type abundance between perturbed and control (x-axis). Each point represents a unique cell type, perturbation pair, and plots are faceted by timepoint. Cell type specific, differentially expressed genes resultant to perturbation pairs are not associated with changes in cell type abundance.