Fig. 5: Inferring epigenome and transcriptome consistent dynamics in mouse hair follicle development using GraphVelo multi-omics velocities.

a GraphVelo velocity fields of mouse hair follicle development. Cells were colored by cell macrostates. b Number of terminal states predicted by CellRank using velocities inferred with different methods. c Driver genes along multiple lineages identified through CellRank. d Topological analyses of GraphVelo vector field identified novel root cells and attractors residing in three terminal states(IRS, hair shaft-cuticle cortex, and medulla). e Expression levels of marker genes in novel root cells and expected root cells. Markers identified by Ma et al.63 were highlighted with stars, and newly identified markers were highlighted in bold. f Regression results of MSD values along the transition path from the expected root or novel root to IRS. Two genes Runx1 and Shh genes with large MSD originating from the novel root were highlighted. The solid line denotes mean of regression result, Shaded region represents 1 s.d. g DTW distance between RNA velocity and chromatin velocity of individual genes. CCD genes were colored in red. The dotted line indicates the elbow point separating the decoupled genes from the rest. h GO enrichment of decoupled genes in (g). i Line plot of nomarlized RNA and chromatin velocity along pseudotime for genes predicted by GraphVelo to have notable decoupling patterns. Chromatin velocity trends were colored as brown and RNA velocity trends were colored as green. j Heatmaps of Jacobian element distribution along the axis of regulator RNA abundance of four regulator effector circuits: i) Lef1 versus Wnt3 chromatin accessibilities. ii) Hoxc13 versus Wnt3 chromatin accessibilities. iii) Lef1 versus Wnt3 transcription. iv) Hoxc13 versus Wnt3 transcription. k Effective dose-response curves obtained from integrating the averaged Jacobian elements over the corresponding normalized regulator mature mRNA regulator level in (j). Source data are provided as a Source Data file.