1）简介：

DESeq2-package： for differential analysis of count data（对count data 做差异分析）

2）安装

if("DESeq2" %in% rownames(installed.packages()) == FALSE) {source("http://bioconductor.org/biocLite.R");biocLite("DESeq2")}
suppressMessages(library(DESeq2))
ls('package:DESeq2')

 3）对象的使用说明

3.1）coef（Extract a matrix of model coefficients/standard errors，高级用户检验模型系数）

语法：coef(object, SE = FALSE, ...)

参数解释：

object：a DESeqDataSet returned by DESeq, nbinomWaldTest, or nbinomLRT.

例子：

dds <- makeExampleDESeqDataSet(m=4)
dds <- DESeq(dds)
coef(dds)[1,]
coef(dds, SE=TRUE)[1,]

3.2） collapseReplicates：Collapse technical replicates in a RangedSummarizedExperiment or DESeqDataSet(用于消除技术重复)
用法：collapseReplicates(object, groupby, run, renameCols = TRUE)
参数：

object：A RangedSummarizedExperiment or DESeqDataSet
groupby：a grouping factor, as long as the columns of object，分组因子
run：optional, the names of each unique column in object. if provided, a new column runsCollapsed will be added to the colData which pastes together the names of run （测序run）
renameCols：whether to rename the columns of the returned object using the levels of the grouping factor

例子：

dds <- makeExampleDESeqDataSet(m=12)
str(dds)
dds$sample <- factor(sample(paste0("sample",rep(1:9, c(2,1,1,2,1,1,2,1,1))))) （#共9个样品：其中 3个样品有2个技术重重）
dds$run <- paste0("run",1:12) #12个run道
ddsColl <- collapseReplicates(dds, dds$sample, dds$run)
# examine the colData and column names of the collapsed data
colData(ddsColl)
colnames(ddsColl)
# check that the sum of the counts for "sample1" is the same
# as the counts in the "sample1" column in ddsColl
matchFirstLevel <- dds$sample == levels(dds$sample)[1]
stopifnot(all(rowSums(counts(dds[,matchFirstLevel])) == counts(ddsColl[,1])))

3.3）counts：Accessors for the ’counts’ slot of a DESeqDataSet object（对表达矩阵进行统计，）

one row for each observational unit (gene or the like), and one column for each sample(行代表观察值(例如基因)，列代表样本(例如肝、脾、肾等))

 语法:counts(object, normalized = FALSE,replaced = FALSE)

 参数:

object:a DESeqDataSet object(表达矩阵).
normalized：logical indicating whether or not to divide the counts by the size factors or normalization factors before returning (normalization factors always preempt size factors)，(即不同量级的数据要不要归一化)
replaced：返回极端值

dds <- makeExampleDESeqDataSet(m=4)  ##构建一个表达矩阵
head(counts(dds))
dds <- estimateSizeFactors(dds) # run this or DESeq() first  
head(counts(dds, normalized=TRUE))

3.4）DESeq：Differential expression analysis based on the Negative Binomial (a.k.a.Gamma-Poisson) distribution（基于负二项分布进行差异分析）

语法：

DESeq(object, test = c("Wald", "LRT"), fitType = c("parametric", "local","mean"), sfType = c("ratio", "poscounts", "iterate"), betaPrior,full = design(object), reduced, quiet = FALSE,minReplicatesForReplace = 7, modelMatrixType, useT = FALSE, minmu = 0.5,
parallel = FALSE, BPPARAM = bpparam())

参数：

object：a DESeqDataSet object（表达矩阵对象）
test：Wald" or "LRT"检验
fitType：either "parametric", "local", or "mean"
sfType：either "ratio", "poscounts", or "iterate" for teh type of size factor estimation.
betaPrior：whether or not to put a zero-mean normal prior on the non-intercept coefficients
reduced：for test="LRT", a reduced formula to compare against
quiet：whether to print messages at each step
minReplicatesForReplace：the minimum number of replicates required
modelMatrixType：either "standard" or "expanded", which describe how the model matrix, X of the GLM formula is formed.
useT：logical, passed to nbinomWaldTest, default is FALSE
minmu：lower bound on the estimated count for fitting gene-wise dispersion
parallel：if FALSE, no parallelization. if TRUE, parallel execution using BiocParallel,
BPPARAM：an optional parameter object passed internally to bplapply when parallel=TRUE.
例子：

# count tables from RNA-Seq data
cnts <- matrix(rnbinom(n=1000, mu=100, size=1/0.5), ncol=10)
cond <- factor(rep(1:2, each=5))

# object construction
dds <- DESeqDataSetFromMatrix(cnts, DataFrame(cond), ~ cond)

# standard analysis
dds <- DESeq(dds)
res <- results(dds)

# moderated log2 fold changes
resultsNames(dds)
resLFC <- lfcShrink(dds, coef=2, type="apeglm")

# an alternate analysis: likelihood ratio test
ddsLRT <- DESeq(dds, test="LRT", reduced= ~ 1)
resLRT <- results(ddsLRT)