Fig. 3: TF perturbations alter EN/IN output of human cortical RG.
From: Dissecting gene regulatory networks governing human cortical cell fate
a, Neighbourhood graphs based on the UMAP in Fig. 1d showing results of differential abundance testing using Milo under perturbation of eight prioritized TFs at the gene level. NT cells were downsampled to match perturbed cell numbers. Node sizes denote cell numbers in each neighbourhood; edges depict the number of cells shared between neighbourhoods; node colour denotes log2FC (false discovery rate (FDR) = 0.1) when perturbed. b, Top, schematics illustrating RG of distinct fate biases; bottom, experimental design for combined TF perturbation and lineage tracing in cortical RG of nine human individuals. c, UMAP of lineage-resolved 129,003 human cells collected on differentiation day 7, coloured by supervised cell type. A median of 15,469 cells per TF perturbation were recovered after filtering. d, Upset plot of cell class compositions in clones containing cells from more than one class (multi-class clones). e, Heatmap for lineage coupling score matrix in main cell types. f, UMAPs showing cell fate distributions in clonal clusters identified using scLiTr89: RG-, EN-, EN/IN-, IN- and oligodendrocyte-biased clusters. g, Box plots showing distributions of clonal cluster fractions in eight human individuals in NR2E1, ARX and ZNF219 KD versus NT. Two-sided paired Wilcoxon tests; P = 0.0078, 0.64, 0.016 (NR2E1); 0.0078, 0.0078, 1 (ARX) and 0.38, 0.023, 0.0078 (ZNF219). Centre line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; points, outliers. *P < 0.05; **P < 0.01; ***P < 0.001. Panel b created in BioRender. Nowakowski, T. (https://BioRender.com/7cr6o12).
