Fig. 5: MuTrans can reveal the underlying complex dynamics in single-cell blood differentiation datasets. | Nature Communications

Fig. 5: MuTrans can reveal the underlying complex dynamics in single-cell blood differentiation datasets.

From: Dissecting transition cells from single-cell transcriptome data through multiscale stochastic dynamics

Fig. 5

a, b The constructed dynamical manifold by MuTrans is shown for the two datasets. The color of each individual cell in dynamical manifold is based on its soft-clustering membership value. In mouse blood cells dataset (a), MuTrans highlights the multi-lineage cells in a shallow pit on dynamical manifold. In the blood cells dataset toward lymphoid lineages (b), MuTrans discovers plenty of transition cells exist between stable Pro-B and Pre-B cell attractors (marked by dashed squares). pDC: plasmacytoid dendritic cells. ILC: innate lymphoid cells. NKC: natural-killer cells. The complete cell type annotations are shown in Supplementary Fig. 12. c, d The TCS (transition cell score) of transition and average gene expression of the top 5 TD (transition-driver, green), MS (meta-stable, blue) and IH (intermediate-hybrid, yellow) genes for the two interested transition paths marked with dash in (a, b). Transition cells are marked by dashed squares. The thin lines represent the raw normalized expression value and thick lines denote the smoothed data. The complete gene lists are shown in Supplementary Table 6–8. Multi-lin: multi-lineage cells. Gran: granulocytic cells.

Back to article page