quickPseudotime: Quick MST-based pseudotime
In TSCAN: Tools for Single-Cell Analysis

Description Usage Arguments Details Value Author(s) See Also Examples

A convenience wrapper to quickly compute a minimum spanning tree (MST) on the cluster centroids to obtain a pseudotime ordering of the cells.

quickPseudotime(x, ...)

## S4 method for signature 'ANY'
quickPseudotime(x, clusters, others = NULL, ..., start = NULL, columns = NULL)

## S4 method for signature 'SummarizedExperiment'
quickPseudotime(x, ..., assay.type = "logcounts")

## S4 method for signature 'SingleCellExperiment'
quickPseudotime(
  x,
  clusters = colLabels(x, onAbsence = "error"),
  ...,
  others = NULL,
  use.dimred = NULL,
  other.dimreds = TRUE
)

`x`	A numeric matrix of coordinates where each row represents a cell/sample and each column represents a dimension (usually a PC or another low-dimensional embedding, but features or genes can also be used). Alternatively, a SummarizedExperiment or SingleCellExperiment object containing such a matrix in its `assays`, as specified by `assay.type`. This will be transposed prior to use. Alternatively, for SingleCellExperiments, this matrix may be extracted from its `reducedDims`, based on the `use.dimred` specification. In this case, no transposition is performed.
`...`	For the generic, further arguments to pass to the specific methods. For the ANY method, further arguments to pass to `createClusterMST`. For the SummarizedExperiment method, further arguments to pass to the ANY method. For the SingleCellExperiment method, further arguments to pass to the SummarizedExperiment method (if `use.dimred` is specified) or the ANY method (otherwise).
`clusters`	A vector or factor of length equal to the number of cells in `x`, specifying the cluster assignment for each cell.
`others`	List of numeric matrices with the same number of rows as `x`, to be passed to `reportEdges`. This typically contains dimensionality reduction results, for use in visualizing the edges of the MST. If `NULL`, defaults to a list containing `x`.
`start`	Arguments passed to `orderCells`.
`columns`	A character, logical or integer vector specifying the columns of `x` to use. If `NULL`, all provided columns are used by default.
`assay.type`	An integer or string specifying the assay to use from a SummarizedExperiment `x`.
`use.dimred`	An integer or string specifying the reduced dimensions to use from a SingleCellExperiment `x`.
`other.dimreds`	Logical scalar indicating whether all dimensionality reduction results in `x` should be appended onto the `others` list.

This function simply calls, in order:

rowmean, to compute the average low-dimensional coordinates for each cluster.
createClusterMST on the average coordinates created from x.
reportEdges on the average coordinates for all entries of other.
mapCellsToEdges on the per-cell coordinates in x with the constructed MST.
orderCells on the mappings generated from x onto the MST.

A List containing:

centered, a list of numeric matrices containing the averaged coordinates for each cluster. Each matrix corresponds to a dimensionality reduction result in x.
mst, a graph object containing the cluster-level MST computed on the coordinates from use.
ordering, a numeric matrix of pseudotimes for various paths through the MST computed from use.
connected, a list of data.frames containing the edge coordinates between centers. Each data.frame corresponds to a dimensionality reduction result in x.

Aaron Lun

createClusterMST and friends, for the functions that do the actual work.

# Mocking up an SCE object:
ncells <- 100
u <- matrix(rpois(20000, 5), ncol=ncells)
pca <- matrix(runif(ncells*5), ncells)
tsne <- matrix(rnorm(ncells*2), ncells)

library(SingleCellExperiment)
sce <- SingleCellExperiment(assays=list(counts=u),
    reducedDims=SimpleList(PCA=pca, tSNE=tsne))

# Clustering on our pretend PCA values:
clusters <- kmeans(pca, 3)$cluster

# Quickly computing the pseudotime:
out <- quickPseudotime(sce, clusters, use.dimred="PCA")
out$mst
head(out$ordering)