Single-Cell Pseudotime Trajectory Explorer

Trajectory Parameters

Cell count 260

Expression noise 0.30

Branch divergence 70°

Bifurcation point 0.40

Live Readouts

Cells

—

Branch points

—

Trajectory depth

—

Selected cell

click a cell

How it works

Each dot is a simulated single cell placed in a 3D expression-space embedding (the low-dimensional stand-in for a real scRNA-seq UMAP). Cells are generated along a branching lineage: a trunk phase, then a bifurcation into two daughter fates, with random noise added to mimic biological and technical variability.

A minimum spanning tree (MST) is built over all cells by Euclidean distance in the embedding — the same principle used by trajectory tools like Monocle and Slingshot to connect cells into a lineage graph without assuming the branching structure in advance:

MST = argmin over spanning trees T of Σ (u,v)∈T  d(u, v)

pseudotime(x) = Σ w(e)  over edges e on the unique path
                root → x  inside the MST

Because the graph is a tree, there is exactly one path between the root and any cell, so its length is unambiguous — that sum is the cell's pseudotime, a proxy for how far it has progressed along differentiation. Nodes where the tree has three or more neighbours are the algorithm's own detected branch points, found purely from graph topology.

Sliders — reshape the underlying biology: more cells, more noise, a wider fork angle, or an earlier/later bifurcation.
Click a cell — selects it and shows its current pseudotime.
Set selected cell as root — recomputes every pseudotime as tree-distance from that cell, without regenerating the data, exactly as choosing a different root in Slingshot re-orders the same lineage graph.

Root cell (pseudotime 0) MST lineage edges

earlypseudotimelate

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.