AllelicImbalance (version 1.6.2)

ASEset-class: ASEset objects

Description

Object that holds allele counts, genomic positions and map-bias for a set of SNPs

Usage

alleleCounts(x, strand = "*", return.class = "list")
"alleleCounts"(x, strand = "*", return.class = "list")
alleleCounts(x, strand = "*") <- value
"alleleCounts"(x, strand = "*") <- value
mapBias(x, ...)
"mapBias"(x, return.class = "list")
fraction(x, ...)
"fraction"(x, strand = "*", top.fraction.criteria = "maxcount", verbose = FALSE)
arank(x, return.type = "names", return.class = "list", strand = "*", ...)
frequency(x, ...)
genotype(x, ...)
"genotype"(x, return.class = "matrix")
genotype(x) <- value
"genotype"(x) <- value
countsPerSnp(x, ...)
"countsPerSnp"(x, return.class = "matrix", return.type = "mean", strand = "*")
countsPerSample(x, ...)
"countsPerSample"(x, return.class = "matrix", return.type = "mean", strand = "*")
phase(x, ...)
"phase"(x, return.class = "matrix")
phase(x) <- value
"phase"(x) <- value
ref(x)
"ref"(x)
ref(x) <- value
"ref"(x) <- value
alt(x)
"alt"(x)
alt(x) <- value
"alt"(x) <- value

Arguments

x

ASEset object

strand

which strand of '+', '-' or '*'

return.class

return 'list' or 'array'

value

replacement variable

...

additional arguments

top.fraction.criteria

'maxcount', 'ref' or 'phase'

verbose

makes function more talkative

return.type

return 'names', rank or 'counts'

Value

An object of class ASEset containing location information and allele counts for a number of SNPs measured in a number of samples on various strand, as well as mapBias information. All data is stored in a manner similar to the SummarizedExperiment class.

Table

table(...)

Arguments:

...

An ASEset object that contains the variants of interest

The generics for table does not easily allow more than one argument so in respect to the different strand options, table will return a SimpleList with length 3, one element for each strand.

Frequency

frequency(x, return.class = "list", strand = "*", threshold.count.sample = 15)

Arguments:

x

An ASEset object that contains the variants of interest

x

threshold.count.samples

if sample has fewer counts the function return NA.

Constructor

ASEsetFromCountList(rowRanges, countListNonStranded = NULL, countListPlus = NULL, countListMinus = NULL, countListUnknown = NULL, colData = NULL, mapBiasExpMean = array(), verbose=FALSE, ...)

Arguments:

rowRanges

A GenomicRanges object that contains the variants of interest

countListNonStranded

A list where each entry is a matrix with allele counts as columns and sample counts as rows

countListPlus

A list where each entry is a matrix with allele counts as columns and sample counts as rows

countListMinus

A list where each entry is a matrix with allele counts as columns and sample counts as rows

countListUnknown

A list where each entry is a matrix with allele counts as columns and sample counts as rows

colData

A DataFrame object containing sample specific data

mapBiasExpMean

A 3D array describing mapping bias. The SNPs are in the 1st dimension, samples in the 2nd dimension and variants in the 3rd dimension.

verbose

Makes function more talkative

...

arguments passed on to SummarizedExperiment constructor

Details

An ASEset object differs from a regular SummarizedExperiment object in that the assays contains an array instead of matrix. This array has ranges on the rows, sampleNames on the columns and variants in the third dimension.

It is possible to use the commands barplot and locationplot on an ASEset object see more details in barplot and locationplot.

Three different alleleCount options are available. The simples one is the * option, and is for experiments where the strand information is not known e.g. non-stranded data. The unknown strand could also be for strand specific data when the aligner could not find any strand associated with the read, but this should normally not happen, and if it does probably having an extremely low mapping quality. Then there are an option too add plus and minus stranded data. When using this, it is essential to make sure that the RNA-seq experiment under analysis has in fact been created so that correct strand information was obtained. The most functions will by default have their strand argument set to '*'.

The phase information is stored by the convention of 'maternal chromosome|paternal chromosome', with 0 as reference allele and 1 as alternative allele. '|' when the phase is known and '/' when the phase is unknown. Internally the information will be stored as an three dimensional array, dim 1 for SNPs, dim 2 for Samples and dim 3 which is fixed and stores maternal chromosome, paternal chromosome and phased (1 equals TRUE).

See Also

• The SummarizedExperiment for ranges operations.

Examples

#make example countList
set.seed(42)
countListPlus <- list()
snps <- c('snp1','snp2','snp3','snp4','snp5')
for(snp in snps){
  count<-matrix(rep(0,16),ncol=4,dimnames=list(
c('sample1','sample2','sample3','sample4'),
c('A','T','G','C')))
  #insert random counts in two of the alleles
  for(allele in sample(c('A','T','G','C'),2)){
count[,allele]<-as.integer(rnorm(4,mean=50,sd=10))
  }
  countListPlus[[snp]] <- count
}

#make example rowRanges
rowRanges <- GRanges(
  seqnames = Rle(c('chr1', 'chr2', 'chr1', 'chr3', 'chr1')),
  ranges = IRanges(1:5, width = 1, names = head(letters,5)),
  snp = paste('snp',1:5,sep='')
)

#make example colData
colData <- DataFrame(Treatment=c('ChIP', 'Input','Input','ChIP'),
 row.names=c('ind1','ind2','ind3','ind4'))

#make ASEset
a <- ASEsetFromCountList(rowRanges, countListPlus=countListPlus,
colData=colData)


#example phase matrix (simple form)
p1 <- matrix(sample(c(1,0),replace=TRUE, size=nrow(a)*ncol(a)),nrow=nrow(a), ncol(a))
p2 <- matrix(sample(c(1,0),replace=TRUE, size=nrow(a)*ncol(a)),nrow=nrow(a), ncol(a))
p <- matrix(paste(p1,sample(c("|","|","/"), size=nrow(a)*ncol(a), replace=TRUE), p2, sep=""),
	nrow=nrow(a), ncol(a))

phase(a) <- p