LouvainDepart: Louvain clustering using departure as data representation

In scpoisson: Single Cell Poisson Probability Paradigm

Louvain clustering using departure as data representation

Description

This function returns a list with elements useful to check and compare cell clustering.

Usage

LouvainDepart(
  data,
  pdat = NULL,
  PCA = TRUE,
  N = 15,
  pres = 0.8,
  tsne = FALSE,
  umap = FALSE,
  ...
)

Arguments

data	A UMI count matrix with genes as rows and cells as columns or an S3 object for class 'scppp'.
pdat	A matrix used as input for cell clustering. If not specify, the departure matrix will be calculated within the function.
PCA	A logic value specifying whether apply PCA before Louvain clustering, default is TRUE.
N	A numeric value specifying the number of principal components included for further clustering (default 15).
pres	A numeric value specifying the resolution parameter in Louvain clustering (default 0.8)
tsne	A logic value specifying whether t-SNE dimension reduction should be applied for visualization.
umap	A logic value specifying whether UMAP dimension reduction should be applied for visualization.
...	not used.

Details

This is a function used to get cell clustering using Louvain clustering algorithm implemented in the Seurat package.

Value

A list with the following elements:

• sdata: a Seurat object

• tsne_data: a matrix containing t-SNE dimension reduction results, with cells as rows, and first two t-SNE dimensions as columns; NULL if tsne = FALSE.

• umap_data: a matrix containing UMAP dimension reduction results, with cells as rows, and first two UMAP dimensions as columns; NULL if tsne = FALSE.

• res_clust: a data frame contains two columns: names (cell names) and clusters (cluster label)

References

\insertRef

Seuratscpoisson

Examples

set.seed(1234)
test_set <- matrix(rpois(500, 2), nrow = 20)
rownames(test_set) <- paste0("gene", 1:nrow(test_set))
colnames(test_set) <- paste0("cell", 1:ncol(test_set))
LouvainDepart(test_set)

scpoisson documentation built on Aug. 17, 2022, 9:05 a.m.