DBclustering: Perform spectral density clustering on single cells

In lyc-1995/MySeuratWrappers: My extentions to Seurat package

Description

Find point clounds single cells in a two-dimensional space using density clustering (DBSCAN).

Usage

## S3 method for class 'Seurat'
DBclustering(
  object,
  assay = NULL,
  reduction = "tsne",
  key = NULL,
  dim.1 = 1,
  dim.2 = 2,
  eps,
  minPts = 5,
  seed.use = 42,
  ...
)

## S3 method for class 'DimReduc'
DBclustering(
  object,
  key = NULL,
  dim.1 = 1,
  dim.2 = 2,
  eps,
  minPts = 5,
  seed.use = 42,
  ...
)

## Default S3 method:
DBclustering(object, eps, minPts = 5, seed.use = 42, ...)

DBclustering(object, ...)

Arguments

object	An object
assay	Assay to use in DBSCAN clustering
reduction	Which dimensional reduction to use, default is tsne.
key	Key used in provided DimReduc object
dim.1	First dimension to use
dim.2	second dimension to use
eps	size of the epsilon neighborhood. Lower value to get more fine-scale clustering.
minPts	number of minimum points in the eps region (for core points). Default is 5 points.
seed.use	Random seed for the dbscan function
...	Arguments passed to other methods and destiny

Value

Returns a Seurat object where the idents have been updated with new cluster info; latest clustering results will be stored in object metadata under 'DB_clusters'. Note that 'DB_clusters' will be overwritten everytime FindClusters is run

Examples

pbmc_small
# Density based clustering on the first two tSNE dimensions
pbmc_small <- DBclustering(pbmc_small)

lyc-1995/MySeuratWrappers documentation built on June 30, 2020, 11:48 a.m.