DoubletDecon: Main DoubletDecon v1.0.1

In lyc-1995/DoubletDeconSeurat: Seurat wrappers with DoubletDecon

Description

This is the main function. This function identifies clusters of doublets with a combination of deconvolution analysis and unique gene expression and individual doublet cells with deconvolution analysis.

Usage

DoubletDecon(
  rawDataFile,
  groupsFile,
  filename,
  location,
  fullDataFile = NULL,
  removeCC = FALSE,
  species = "mmu",
  rhop = 1,
  write = TRUE,
  PMF = TRUE,
  useFull = FALSE,
  heatmap = TRUE,
  centroids = FALSE,
  num_doubs = 100,
  only50 = FALSE,
  min_uniq = 4,
  seed = 100
)

Arguments

rawDataFile	Name of file containing ICGS or Seurat expression data (gene by cell)
groupsFile	Name of file containing group assignments (3 column: cell, group(numeric), group(numeric or character))
filename	Unique filename to be incorporated into the names of outputs from the functions.
location	Directory where output should be stored
fullDataFile	Name of file containing full expression data (gene by cell). Default is NULL.
removeCC	Remove cell cycle gene cluster by KEGG enrichment. Default is FALSE.
species	Species as scientific species name, KEGG ID, three letter species abbreviation, or NCBI ID. Default is "mmu".
rhop	x in mean+x*SD to determine upper cutoff for correlation in the blacklist. Default is 1.
write	Write output files as .txt files. Default is TRUE.
PMF	Use step 2 (unique gene expression) in doublet determination criteria. Default is TRUE.
useFull	Use full gene list for PMF analysis. Requires fullDataFile. Default is FALSE.
heatmap	Boolean value for whether to generate heatmaps. Default is TRUE. Can be slow to datasets larger than ~3000 cells.
centroids	Use centroids as references in deconvolution instead of the default medoids.
num_doubs	The user defined number of doublets to make for each pair of clusters. Default is 100.
only50	use only synthetic doublets created with 50%/50% mix of parent cells, as opposed to the extended option of 30%/70% and 70%/30%, default is FALSE.
min_uniq	minimum number of unique genes required for a cluster to be rescued
seed	Set a random seed.

Value

data_processed = new expression file (cleaned).

groups_processed = new groups file (cleaned).

PMF_results = pseudo marker finder t-test results (gene by cluster).

DRS_doublet_table = each cell and whether it is called a doublet by deconvolution analysis.

DRS_results = results of deconvolution analysis (cell by cluster) in percentages.

Decon_called_freq = percentage of doublets called in each cluster by deconvolution analysis.

Final_doublets_groups = new groups file containing only doublets.

Final_nondoublets_groups = new groups file containing only non doublets.

lyc-1995/DoubletDeconSeurat documentation built on March 8, 2020, 5:47 a.m.