runPCA: runPCA

In bernard2012/Giotto: Spatial single-cell transcriptomics toolbox

Description

runs a Principal Component Analysis

Usage

runPCA(
  gobject,
  expression_values = c("normalized", "scaled", "custom"),
  reduction = c("cells", "genes"),
  name = "pca",
  genes_to_use = "hvg",
  return_gobject = TRUE,
  center = F,
  scale_unit = F,
  ncp = 100,
  method = c("irlba", "factominer"),
  rev = FALSE,
  verbose = TRUE,
  ...
)

Arguments

gobject	giotto object
expression_values	expression values to use
reduction	cells or genes
name	arbitrary name for PCA run
genes_to_use	subset of genes to use for PCA
return_gobject	boolean: return giotto object (default = TRUE)
center	center data first (default = FALSE)
scale_unit	scale features before PCA (default = FALSE)
ncp	number of principal components to calculate
method	which implementation to use
rev	do a reverse PCA
verbose	verbosity of the function
...	additional parameters for PCA (see details)

Details

See prcomp_irlba and PCA for more information about other parameters.

• genes_to_use = NULL: will use all genes from the selected matrix

• genes_to_use = <hvg name>: can be used to select a column name of highly variable genes, created by (see calculateHVG)

• genes_to_use = c('geneA', 'geneB', ...): will use all manually provided genes

Value

giotto object with updated PCA dimension recuction

Examples

# 1. create giotto object
expr_path = system.file("extdata", "seqfish_field_expr.txt", package = 'Giotto')
loc_path = system.file("extdata", "seqfish_field_locs.txt", package = 'Giotto')
VC_small <- createGiottoObject(raw_exprs = expr_path, spatial_locs = loc_path)

# 2. normalize giotto
VC_small <- normalizeGiotto(gobject = VC_small, scalefactor = 6000)
VC_small <- addStatistics(gobject = VC_small)

# 3. dimension reduction
VC_small <- calculateHVG(gobject = VC_small)
VC_small <- runPCA(gobject = VC_small)
plotPCA(VC_small)

bernard2012/Giotto documentation built on Sept. 22, 2020, 10:29 a.m.