Skip to contents

Models independent developmental trajectories for an arbitrary number of genes. For each gene, it isolates the "Core" expressing cells to calculate an expression-weighted barycenter and covariance matrix, bypassing the gravitational pull of ambient RNA noise. It then draws a perfect geometric ellipse using the Chi-Square distribution. Cells falling within the confidence interval of a gene's trajectory are marked as TRUE for that specific tube, allowing for overlapping dual identities.

Usage

mExtractTrajectoryTubes(object, ...)

# S4 method for class 'mCanonicalTockyObj'
mExtractTrajectoryTubes(
  object,
  seurat_obj,
  features,
  ci = 0.9,
  window_size = 10,
  span = 0.4,
  reset = TRUE
)

Arguments

object

An mCanonicalTockyObj.

...

Additional arguments passed to methods.

seurat_obj

Original Seurat object.

features

Character vector of genes to model trajectories for (e.g., c("Cd4", "Cd8a")).

ci

Numeric. Confidence interval for the ellipses (default 0.90).

window_size

Numeric. Width of the sliding angle slice (default 10).

span

Numeric. LOESS smoothing span for the barycenters (default 0.4).

reset

Logical. If TRUE (default), deletes previously created lineage tubes. If FALSE, appends new tubes and updates the composite Lineage_Identity.