Fig. 6: VIA accurately infers global connectivity and sub-trajectories in the 1.3-million scRNA-seq mouse organogenesis cell atlas. | Nature Communications

Fig. 6: VIA accurately infers global connectivity and sub-trajectories in the 1.3-million scRNA-seq mouse organogenesis cell atlas.

From: Generalized and scalable trajectory inference in single-cell omics data with VIA

Fig. 6: VIA accurately infers global connectivity and sub-trajectories in the 1.3-million scRNA-seq mouse organogenesis cell atlas.

a MOCA graph trajectory (nodes colored by pseudotime) and shaded-colored regions corresponding to major cell groups. Stem branch consists of epithelial cells derived from ectoderm and endoderm, leading to two main branches: (1) the mesenchymal and (2) the neural tube and neural crest. Other major groups are placed in the biologically relevant neighborhoods, such as the adjacencies between hepatocyte and epithelial trajectories; the neural crest and the neural tube; as well as the links between early mesenchyme with both the hematopoietic cells and the endothelial cells (see Supplementary Note 7). b Colored by VIA pseudotime. c Lineage pathways and probabilities of neuronal, myocyte and WBC lineages. d VIA graph preserves key relationships across choice of number of PCs, whereas (e) UMAP embedding is first step in Monocle3 and highly susceptible to choice of number of PCs (or K in KNN see Fig. 22).

Back to article page