scUnifrac: scUnifrac

In liuqivandy/scUnifrac: a package for quantitative assessment of cell population diversity in single-cell landscapes

Description

Quantify cell population diversity between two single cell RNA-seq datasets

Usage

scUnifrac(data1, sampleName1 = "S1", data2, sampleName2 = "S2",
  ref.expr = NULL, genenum = 500, ncluster = 10, nDim = 4,
  normalize = T, report = T, outputFile = "scUnifrac_report.html")

Arguments

data1	matrix; the data matrix of the first dataset, row is the gene symbol, the column is the cell id
sampleName1	a character string giving the name of the first dataset; (default: "S1")
data2	matrix; the data matrix of the second dataset, row is the gene symbol, the column is the cell id; data1 and data1 should have the same gene symbols
sampleName2	a character string giving the name of the second dataset; (default: "S2")
ref.expr	matrix; the data matrix of the reference cell atlas; the ref.expr is used to predict the cell types of single cell from data1 and data2 (default: NULL)
genenum	integer; Number of highly variable genes to build the cell population structure (default: 500)
ncluster	integer; Number of clusters to divide cells (default: 10)
nDim	integer; Number of PCA dimensions to build the cell population structure (default: 4)
normalize	logical; Indicate whether normalize data1 and data2 (default: TRUE, normalize to the total count and log2 transform)
report	logical; Indicate whether to generate a report (default: TRUE); if set to FALSE, scUnifrac will not generate the report but calculate the distance and the pvalue; set to FALSE, when users have more than two samples to compare and only want to calculate the pairwise distance
outputFile	a character string giving the name of the report;(default: "scUnifrac_report.html")

Value

List with the following elements:

distance	The distance of cell population diversity between two single-cell RNA-seq datasets
pvalue	The statistical signficance of the distance

Examples

library(scUnifrac)  
##load the two example datasets 
load(system.file("extdata", "colon1.Rdata", package = "scUnifrac"))
load(system.file("extdata", "pan1.Rdata", package = "scUnifrac"))
##Calculate the distance and pvalue between data1 and data2 and generate a report summarizing the result in the work directory
result<-scUnifrac( data1=colon1, data2=pan1)  
result
##load the mouse cell altas from Han et al., 2018, Cell 172, 1091–1107. The atlas is used as a reference to predict cell types of data1 and data2
load(system.file("extdata", "ref.expr.Rdata", package = "scUnifrac"))
##Generate a report which also includes the predicted cell types by mapping each cell to the reference 
result<-scUnifrac( data1=colon1, data2=pan1,ref.expr=ref.expr, outputFile="scUnifrac_report.html") 

##test two samples with similar populations
ind<-sample(c(1:ncol(colon1)), ncol(colon1)/2)
result<-scUnifrac(data1=colon1[,ind], data2=colon1[,-ind],ref.expr=ref.expr, outputFile="scUnifrac_report.html")

liuqivandy/scUnifrac documentation built on Jan. 21, 2021, 2:02 p.m.