testDATs: Run differential abundance testings

In EDDA: Experimental Design in Differential Abundance analysis

Description

Perform differential abundance testing on simulated count data.

Usage

testDATs(data, numCores=10, minCountsThreshold=0,
	DE.methods=c("Cuffdiff","DESeq","baySeq","edgeR","MetaStats","NOISeq"), 
	nor.methods=c("default","Mode","UQN","NDE"),method.list=NULL)

Arguments

data	Data object from generateData() function or predifined data object similar to the output of generateData().
numCores	Number of cores for parallelization. Default is 10.
minCountsThreshold	Minimum counts threshold for filtering. Default is 0.
DE.methods	Method list for differential expression tests. Methods currently available include "Cuffdiff","DESeq","baySeq","edgeR","MetaStats","NOISeq".
nor.methods	Normalization method list. Methods currently available include "default" (default normalization for each DE method),"Mode"(Mode normalization), "UQN"(Upper quartile normalization), "NDE"(non-differential expression normalization).
method.list	The method list for the combination of DE.methods and nor.methods. Default is NULL.

Value

data	Data object from generateData() function.
filterCounts	filtered count data.
Cuffdiff	Result form Cuffdiff with default normalization.
Cuffdiff_uqn	Result form Cuffdiff with Upper quartile normalization normalization.
Cuffdiff_Mode	Result form Cuffdiff with Mode normalization.
Cuffdiff_nde	Result form Cuffdiff with non-differential expression normalization.
DESeq	Result form DESeq with default normalization.
DESeq_uqn	Result form DESeq with Upper quartile normalization normalization.
DESeq_Mode	Result form DESeq with Mode normalization.
DESeq_nde	Result form DESeq with non-differential expression normalization.
baySeq	Result form baySeq with default normalization.
baySeq_uqn	Result form baySeq with Upper quartile normalization normalization.
baySeq_Mode	Result form baySeq with Mode normalization.
baySeq_nde	Result form baySeq with non-differential expression normalization.
edgeR	Result form edgeR with default normalization.
edgeR_uqn	Result form edgeR with Upper quartile normalization normalization.
edgeR_Mode	Result form edgeR with Mode normalization.
edgeR_nde	Result form edgeR with non-differential expression normalization.
MetaStats	Result form MetaStats with default normalization.
MetaStats_uqn	Result form MetaStats with Upper quartile normalization normalization.
MetaStats_Mode	Result form MetaStats with Mode normalization.
MetaStats_nde	Result form MetaStats with non-differential expression normalization.
NOISeq	Result form NOISeq with default normalization.
NOISeq_uqn	Result form NOISeq with Upper quartile normalization normalization.
NOISeq_Mode	Result form NOISeq with Mode normalization.
NOISeq_nde	Result form NOISeq with non-differential expression normalization.

Author(s)

Li Juntao, Luo Huaien, Chia Kuan Hui Burton, Niranjan Nagarajan

References

Luo Huaien, Li Juntao,Chia Kuan Hui Burton, Shyam Prabhakar, Paul Robson, Niranjan Nagarajan, The importance of study design for detecting differentially abundant features in high-throughput experiments, under review.

Examples

data <- generateData(EntityCount=100)
test.obj <- testDATs(data,nor.methods="default")
test.obj <- testDATs(data,DE.methods="DESeq")


# test data with input count.
x <- matrix(rnbinom(1000*15,size=1,mu=10), nrow=1000, ncol=15);
x[1:50,11:15] <- x[1:50,11:15]*10
x.name=paste("g",1:1000,sep="");
write.table(cbind(x.name,x),"count.txt",row.names =FALSE, sep ='\t')

x <- read.table("count.txt",head=TRUE,sep='\t')
x.count <- x[,2:16]
x.lable=c(rep(0,10),rep(1,5))
row.names(x.count) <- x[,1]
data <- list(count=x.count,dataLabel=x.lable)
test.obj <- testDATs(data,DE.methods=c("DESeq","edgeR"),nor.methods="default")

EDDA documentation built on Nov. 8, 2020, 5:44 p.m.