runSeuratUMAP: runSeuratUMAP Computes UMAP from the given sce object and...

View source: R/seuratFunctions.R

runSeuratUMAPR Documentation

runSeuratUMAP Computes UMAP from the given sce object and stores the UMAP computations back into the sce object

Description

runSeuratUMAP Computes UMAP from the given sce object and stores the UMAP computations back into the sce object

Usage

runSeuratUMAP(
  inSCE,
  useReduction = c("pca", "ica"),
  reducedDimName = "seuratUMAP",
  dims = 10,
  minDist = 0.3,
  nNeighbors = 30L,
  spread = 1,
  externalReduction = NULL,
  seed = 42,
  verbose = TRUE
)

Arguments

inSCE

(sce) object on which to compute the UMAP

useReduction

Reduction to use for computing UMAP. One of "pca" or "ica". Default is "pca".

reducedDimName

Name of new reducedDims object containing Seurat UMAP Default seuratUMAP.

dims

Numerical value of how many reduction components to use for UMAP computation. Default 10.

minDist

Sets the "min.dist" parameter to the underlying UMAP call. See RunUMAP for more information. Default 0.3.

nNeighbors

Sets the "n.neighbors" parameter to the underlying UMAP call. See RunUMAP for more information. Default 30L.

spread

Sets the "spread" parameter to the underlying UMAP call. See RunUMAP for more information. Default 1.

externalReduction

Pass DimReduc object if PCA/ICA computed through other libraries. Default NULL.

seed

Random seed for reproducibility of results. Default 42.

verbose

Logical value indicating if informative messages should be displayed. Default is TRUE.

Value

Updated sce object with UMAP computations stored

Examples

data(scExample, package = "singleCellTK")
## Not run: 
sce <- runSeuratNormalizeData(sce, useAssay = "counts")
sce <- runSeuratFindHVG(sce, useAssay = "counts")
sce <- runSeuratScaleData(sce, useAssay = "counts")
sce <- runSeuratPCA(sce, useAssay = "counts")
sce <- runSeuratFindClusters(sce, useAssay = "counts")
sce <- runSeuratUMAP(sce, useReduction = "pca")

## End(Not run)

compbiomed/singleCellTK documentation built on Oct. 27, 2024, 3:26 a.m.