Extended Data Fig. 9: Pseudolineage analysis and neurogenesis.

a, Pseudolineage tree algorithm, computing shortest paths to the root cell passing only through neighbours inside the information radius (that is, neighbours within a fixed maximum Jensen–Shannon distance). b, Isolating the pseudolineages terminating in a selected cluster, and projecting to the pseudoage axis. c, Ten selected pseudolineages coloured by major class, on the t-SNE of Fig. 1 with grey scale showing the geodesic distance to the root from every cell. d, The same ten selected pseudolineages coloured by tissue. e–i, Pseudolineages of astrocytes (e), OPCs (f), fibroblasts (g), cortical excitatory neurons (h), forebrain GABAergic neurons (i), midbrain dopaminergic neurons (j) and hindbrain serotonergic neurons (k). Each plot shows a randomly selected subset of pseduolineages terminating in the indicated clusters, as well as expression of selected genes in pseudoage bins along the lineage (calculated for all cells in the lineage). The region and class of each cell is indicated at the bottom. l, Expression of Nhlh1 and Nhlh2. m, In situ hybridization for Nhlh1 and Nhlh2 on sagittal sections of the E115 mouse embryo (Allen Brain Atlas). n, RNA in situ hybridization of mouse embryonic brain at the indicated time points, showing genes relevant to the cortical lineage (image credit: Allen Institute). Each subpanel shows a strip from ventricular zone to pia at four different ages.