EvaluateIdentKmeans: Evaluate the cell identities using K-means clustering

In ncchung/SeuratAddon: Unsupervised evaluation of cell identities in Seurat

Description

Evaluate the cell identities that were determined from k-means clustering

Usage

EvaluateIdentKmeans(object, prob.use = "pip", p.adjust.methods = "BH",
  genes.use = NULL, reduction.type = NULL, dims.use = NULL,
  center = TRUE, data.cut = NULL, minibatch = FALSE,
  assay.type = "RNA", do.plot = TRUE, seed = 1, verbose = TRUE,
  return.jackstraw = FALSE, ...)

Arguments

object	Seurat object
prob.use	Probabilities saved to [e-mail]["ident_prob"]; To be used in downstream analysis by default
p.adjust.methods	A p-value correction method for multiple comparisons
genes.use	Genes to use for clustering
reduction.type	Name of dimensional reduction technique (e.g. "pca", "ica")
dims.use	A vector of the dimensions to use (e.g. To use the first 10 PCs, pass 1:10)
center	Center the cells
data.cut	Clip all z-scores to have an absolute value below this. Reduces the effect of huge outliers in the data. (default is NULL)
minibatch	FALSE by default. If TRUE, use the mini-batch K-means clustering implemented in the ClusterR package.
assay.type	Type of assay to fetch data for (default is RNA)
do.plot	Draw histograms of statistics (default is FALSE)
seed	Random seed
verbose	Print the computational progress. TRUE, by default.
return.jackstraw	FALSE by default. If TRUE, return a list of jackstraw-related statistics, instead of a Seurat object,
...	Additional parameters passed to jackstraw.kmeans

Details

Use data.cut to trim outliers, as done in DoKMeans2. Ensure that these parameters are identical to the original clustering step.

Value

Seurat object where the unsupervised evaluation for cell identities is stored in object@ident_prob. Furthermore, chosen probabilities (PIP, adjusted p-values, or etc) are stored in [e-mail]["ident_prob"]. See the optional argument prob.use.

See Also

DoKMeans2

Examples

## Not run: 
set.seed(1234)
require(Seurat)
  pbmc_small <- DoKMeans2(pbmc_small, k.cells = 3)
  pbmc_small2 <- EvaluateIdentKmeans(pbmc_small)

## End(Not run)

ncchung/SeuratAddon documentation built on May 3, 2019, 3:17 p.m.