add_dimred: Add or create a dimensionality reduction

In dynwrap: Representing and Inferring Single-Cell Trajectories

Add or create a dimensionality reduction

Description

This can also perform dimensionality reduction of

• The projected expression state with RNA velocity, only if dimred is a function and pair_with_velocity=TRUE

• The trajectory, by projecting the milestones and some "waypoints" to the reduced space, only if dataset contains a trajectory

Usage

add_dimred(
  dataset,
  dimred,
  dimred_milestones = NULL,
  dimred_segment_progressions = NULL,
  dimred_segment_points = NULL,
  project_trajectory = TRUE,
  connect_segments = FALSE,
  pair_with_velocity = !is.null(dataset$expression_future),
  expression_source = "expression",
  ...
)

is_wrapper_with_dimred(dataset)

get_dimred(
  dataset,
  dimred = NULL,
  expression_source = "expression",
  return_other_dimreds = FALSE
)

Arguments

dataset	A dataset created by wrap_data() or wrap_expression()
dimred	Can be A function which will perform the dimensionality reduction, see dyndimred::list_dimred_methods() A matrix with the dimensionality reduction, with cells in rows and dimensions (comp_1, comp_2, ...) in columns
dimred_milestones	An optional dimensionality reduction of the milestones. A matrix with milestones in rows and components (comp_1, comp_2, ...) in columns This will be automatically calculated if project_trajectory = TRUE
dimred_segment_progressions, dimred_segment_points	An optional set of points along the trajectory with their dimensionality reduction. dimred_segment_progressions is a dataframe containing the from and to milestones, and their progression. dimred_segment_points is a matrix with points (the same number as in dimred_segment_progressions) in rows and components (comp_1, comp_2, ...) in columns. Both objects have the same number of rows. These will be automatically calculated if project_trajectory = TRUE
project_trajectory	Whether to also project the trajectory. Only relevant if dataset contains a trajectory, and dimred_segment_progressions and dimred_segment_points are not provided
connect_segments	Whether to connect segments between edges
pair_with_velocity	Can perform dimensionality reduction if dimred is a function.
expression_source	The source of expression, can be "counts", "expression", an expression matrix, or another dataset which contains expression
...	extra information to be stored in the wrapper
return_other_dimreds	Whether or not to return also the milestone dimreds and the segment dimreds, if available.

Value

A dataset object with dimred, which is a numeric matrix with cells in rows and the different components in columns.

• If the dataset contained a trajectory, and project_trajectory=TRUE (default), dimred_milestones, dimred_segment_progressions and dimred_segment_points will also be present. These are described in project_trajectory().

See Also

dyndimred::list_dimred_methods(), project_trajectory()

Examples

if (requireNamespace("dyndimred", quietly = TRUE)) {
  dataset <- example_dataset
  dataset <- add_dimred(
    dataset,
    dyndimred::dimred_landmark_mds
  )
  head(dataset$dimred)
}

dynwrap documentation built on July 26, 2023, 5:15 p.m.