Combine_scRNAseq: combine multiple scRNAseq datasets into one and generate QC...

In liuqivandy/scRNABatchQC: a package for quality control of multiple single cell RNAseq data

Description

Combine and compare multiple single cell RNAseq datasets, including
1)combine into one dataset;
2)find highly variable genes, reduce dimensions using PCA and tSNE for the combined dataset;
3)pairwise comparsion between any two datasets to find the top differentially expressed genes;

Usage

Combine_scRNAseq(
  sces,
  nHVGs = 1000,
  nPCs = 10,
  logFC = 1,
  FDR = 0.01,
  sampleRatio = 1,
  organism = "mmusculus"
)

Arguments

sces	a list of SingleCellExperiment objects (results from Process_scRNAseq); each object corresponds to one single cell RNAseq dataset
nHVGs	integer; the number of highly variable genes (default:1000)
nPCs	integer; the number of principal components (default:10)
logFC	float; log fold change cutoff to select differentially expressed genes (default: 1)
FDR	float; FDR cutoff to select differentially expressed genes (default: 0.01)
sampleRatio	float; the ratio of cells sampled from each dataset to examine the expression similarity(default: 1)
organism	string; the organism of single cell RNAseq datasets;if supported by WebGestaltR, the functional enrichment will be performed; (defeault: mmusculus)

Value

a SingleCellExperiment object with several slots:

• assays; ShallowSimpleListAssays object containing one sparse matrix logcounts (log-transformed normalized counts)

• rowRanges@elementMetadata; A Dataframe hvg containining mean, variance and z-score for each gene

• colData; A Dataframe containing conidtion for each cell

• metadata: A list containing other metadata, including

  • reducedDims: a list containing two types of reduced dimensions: PCA and tSNE

  • diffFC: differential expressed genes and pathways from pairwise comparison between any two datasets

  • other metadata including nHVGs,nPC, logFC, FDR, sampleRatio

See Also

Process_scRNAseq , generateReport

Examples

library(scRNABatchQC)
sces<-Process_scRNAseq(inputs=c("https://github.com/liuqivandy/scRNABatchQC/raw/master/bioplar1.csv.gz","https://github.com/liuqivandy/scRNABatchQC/raw/master/bioplar5.csv.gz"))
scesMerge <- Combine_scRNAseq(sces)
logcounts(scesMerge)[1:5,1:5]
head(scesMerge@rowRanges@metadata$hvg)
summary(scesMerge@metadata$reducedDims$PCA)
#visualize PCA results
plot(scesMerge@metadata$reducedDims$PCA$x[,1:2],pch=16,col=as.factor(scesMerge@colData$condition),xlab="PCA1",ylab="PCA2")
#visualize tSNE results
plot(scesMerge@metadata$reducedDims$tSNE,pch=16,col=as.factor(scesMerge@colData$condition),xlab="tSNE1",ylab="tSNE2")
scesMerge@metadata$diffFC

liuqivandy/scRNABatchQC documentation built on March 24, 2021, 11:01 p.m.