Vol. 01 · No. 1 · Beta Edition
Node reference

Trajectory nodes.

Continuous structure in the data: PAGA topology between clusters, diffusion map and DPT pseudotime, a force-directed layout, and RNA velocity from spliced / unspliced counts (scvelo stochastic or dynamical mode) with its velocity graph, latent time and terminal states.

08 nodesCategory · trajectoryin 6 templates

Typical pipelines

Pseudotime: Neighbors → Diffusion Map → DPT (with PAGA alongside) → Trajectory Visualization. Velocity: a loader that keeps spliced / unspliced layers → RNA Velocity → Velocity Graph → Velocity Visualization, and in dynamical mode → Latent Time → Terminal States.

Draw Graph uses ForceAtlas2 when it is available and otherwise falls back to Fruchterman–Reingold. Since v0.25.0 the node reports the layout it actually used (in the stored obsm key and in the Methods text) and the layout is seeded, so the same seed gives the same picture.

Common pitfalls

DPT needs a root. Pick the root cell or root cluster deliberately; pseudotime is 0 at the root and grows away from it.

Velocity needs velocity layers. A plain H5AD without spliced / unspliced layers cannot run RNA Velocity — use Load Velocyto Loom, Load STARsolo Velocity, or an H5AD that already carries them.

Latent Time is dynamical-mode only. Stochastic velocity does not fit the kinetics Latent Time is derived from.

PAGA node

PAGA

paga

Partition-based Graph Abstraction — coarse-grained trajectory topology between clusters.

Node LibraryscRNA-seq
InAnnData (with neighbors + clusters)
OutAnnData (uns["paga"])
Parameters (3)
NameTypeDefaultDescription
groupsobs columnleidenCluster column.
modelselectv1.2"v1.2" (improved) or "v1.0".
use_rna_velocitycheckboxfalseDirected edges from RNA velocity.
DPT node

DPT

dpt

Diffusion Pseudotime — orders cells along a continuous trajectory (0 = root, 1 = differentiated).

Node LibraryscRNA-seq
InAnnData (with neighbors + diffusion map)
OutAnnData (obs["dpt_pseudotime"])
Parameters (6)
NameTypeDefaultDescription
groupsobs columnleidenCluster column.
n_dcsnumber10Diffusion components.
root_grouptext—Root cluster name.
root_cellnumber—Index of root cell (overrides root_group).
n_branchingsnumber0Number of branchings to detect.
min_group_sizenumber0.01Minimum fraction of cells per group.
Diffusion Map node

Diffusion Map

diffusionMap

Diffusion map embedding — captures continuous transitions. Required before DPT.

Node LibraryscRNA-seq
InAnnData (with neighbors)
OutAnnData (obsm["X_diffmap"])
Parameters (1)
NameTypeDefaultDescription
n_compsnumber15Number of diffusion components.
Draw Graph node

Draw Graph

drawGraph

Force-directed graph layout (ForceAtlas2 / Fruchterman-Reingold).

Node LibraryscRNA-seq
InAnnData (with neighbors)
OutAnnData (obsm["X_draw_graph_*"])
Parameters (4)
NameTypeDefaultDescription
layoutselectfafa/fr/drl/kk.
init_posselectpagapaga (recommended)/spectral/random.
n_iterationsnumber—Number of iterations. null = auto.
random_statenumber0Random seed.
RNA Velocity node

RNA Velocity

rnaVelocity

Estimate RNA velocity from spliced/unspliced counts (scvelo).

Node LibraryscRNA-seq
InAnnData (with spliced/unspliced layers)
OutAnnData (layers["velocity"])
Parameters (6)
NameTypeDefaultDescription
modeselectstochasticstochastic/deterministic/dynamical.
n_top_genesnumber2000Top variable genes for velocity.
min_shared_countsnumber20Min shared spliced/unspliced.
n_neighborsnumber30Neighbors for moment computation.
n_pcsnumber30PCA components.
enforcecheckboxfalseForce recomputation even if results exist.
Velocity Graph node

Velocity Graph

velocityGraph

Compute velocity transition graph for stream/arrow plots.

Node LibraryscRNA-seq
InAnnData (with velocity)
OutAnnData (uns["velocity_graph"])
Parameters (3)
NameTypeDefaultDescription
n_neighborsnumber—Neighbors for graph.
approxcheckboxfalseApproximate the velocity graph on a PCA subspace. Off (exact, the scVelo default) is recommended; on is faster but can flip arrow directions.
n_recurse_neighborsnumber—Number of recursive neighbors. null = auto.
Latent Time node

Latent Time

latentTime

Latent time from dynamical RNA velocity. More accurate than DPT when velocity available.

Node LibraryscRNA-seq
InAnnData (with dynamical velocity)
OutAnnData (obs["latent_time"])
Parameters (3)
NameTypeDefaultDescription
min_likelihoodnumber0.1Gene likelihood threshold.
root_keytext—Root cells key (auto-detect if empty).
end_keytext—Key for terminal cells. Empty = auto-detect.
Terminal States node

Terminal States

terminalStates

Identify terminal/initial states from velocity transition matrix.

Node LibraryscRNA-seq
InAnnData (with velocity graph)
OutAnnData (obs["terminal_states"])
Parameters (2)
NameTypeDefaultDescription
groupbyobs column—Cluster column for grouping.
self_transitionscheckboxfalseInclude self-transitions.