Fig. 6: SimiC’s inferred GRNs across timepoints on a regenerating liver (hepatocyte dataset65) capture regulatory dynamics of hepatocytes in different functional states.
From: SimiC enables the inference of complex gene regulatory dynamics across cell phenotypes
a Heatmap of the regulatory dissimilarity score of quiescent, proliferating, metabolically hyperactive, and transitioning cell types across the four sequential timepoints for the different regulons, showing cell and time dependent variation. b Density plots showing the regulatory activity for the regulon YBX1 on the proliferating cell population and for the regulon CEBPB on the quiescent and proliferating cell populations, for the four considered timepoints. Violin plots showing the distribution of the regulatory dissimilarity scores for regulons CEBPA (top) and CEBPB (bottom) across timepoints (c) and across cellular states of the liver regeneration, for the four considered timepoints (d). tSNE plot showing the hepatocyte cells, colored by their regulatory activity score as computed by SimiC (e), SCENIC (f), and SINCERITIES (g) for the regulons CEBPB (top) and CEBPA (bottom). No results for CEBPA with SINCERITIES are shown as this TF did not appear in the inferred GRN. Similarly, ICAnet did not generate any modules with the CEBPA or CEBPB TFs. Timepoint to which each cell belongs is also specified. h Cell state clustering performance as measured by the ARI score. For SCENIC we consider the binarized and non-binarized scores. A larger ARI indicates a better clustering performance.
