Fig. 5
From: DeepVelo: deep learning extends RNA velocity to multi-lineage systems with cell-specific kinetics
Functional enrichment of DeepVelo predicted driver genes. a Overlap of the top 100 driver genes from scVelo and DeepVelo for GABAergic and gliogenic lineages with annotated lineage marker genes. b Ranking density of marker-overlapping driver genes (across all 2000 tested genes) for scVelo and DeepVelo, separated by the GABAergic and gliogenic lineages, respectively. c Overlap of top 100 driver genes from DeepVelo and scVelo for both lineages with annotated transcription factors. d Pathway enrichment analysis results for the top 100 scVelo and DeepVelo driver genes, respectively, in the GABAergic and gliogenic lineages. e Functional signal in the enriched pathways for scVelo and DeepVelo, based on the presence of pathways involved directly in neurogenesis (“Neurogenesis”), not specific to neurogenesis but involved in development (“Developmental non-neuronal”), and not specific to either development or neurogenesis (“Non-specific”). f The top 20 DeepVelo pathway enrichment analysis results, based on FDR corrected p-values, for the GABAergic and gliogenic lineages, respectively