mhMoveK1: Metropolis-Hastings move when number of nucleosomes is equal...

In ArnaudDroitLab/RJMCMC: Bayesian hierarchical model for genome-wide profiling of nucleosome positions based on high-throughput short-read data (MNase-Seq data)

Description

Attempt to randomly change the position of one nucleosome. Special case for a number of nucleosomes equal to one. This move is given by the usual Metropolis-Hastings move, see for instance Tierney (1994).

Usage

mhMoveK1(paramValues, kValue, muValue, sigmafValue, sigmarValue, deltaValue,
  wValue, dfValue, aValue, dimValue)

Arguments

paramValues	a list containing: startPSF a vector of positive integer, the start position of all the forward reads. startPSR a vector of positive integer, the start position of all the reverse reads. kmax a integer the maximum number of nucleosomes allowed. lambda a positive numeric, the theorical mean of the Poisson distribution. minReads a integer, the minimum number of reads in a potential canditate region. y a vector of numeric, the position of all reads ( including forward and reverse reads). nr a integer, the number of reverse reads. nf a integer, the number of forward reads. nbrReads a integer, the total number of reads (including forward and reverse reads). zeta integer, the length of a nucleosome. deltamin a integer, the minimum distance between the maxima of the forward and reverse reads position densities for each nucleosome. deltamax a integer, the maximum distance between the maxima of the forward and reverse reads position densities for each nucleosome. minReadPos a numeric, the minimum position of the reads. maxReadPos a numeric, the maximum position of the reads.
kValue	a integer, the number of nucleosomes.
muValue	a vector of numeric of length kValue, the positions of the nucleosomes.
sigmafValue	a vector of numeric of length kValue, the variance of the forward reads for each nucleosome.
sigmarValue	a vector of numeric of length kValue, the variance of the reverse reads for each nucleosome.
deltaValue	a vector of numeric of length kValue, the distance between the maxima of the forward and reverse reads position densities for each nucleosome.
wValue	a vector of positive numerical of length kValue, the weight for each nucleosome.
dfValue	a vector of positive numerical of length kValue, the degrees of freedom for each nucleosome.
aValue	a vector of positive numerical of length kValue + 1, the positions, on the chromosome, delimiting the regions of the reads associated with each nucleosome.
dimValue	a vector of positive integer of length kValue, the number of reads associated to each nucleosome.

Value

a list containing:

• k a integer, the updated number of nucleosomes.

• mu a vector of numeric of length k, the updated positions of the nucleosomes.

• sigmaf a vector of numeric of length k, the updated variance of the forward reads for each nucleosome.

• sigmar a vector of numeric of length k, the updated variance of the reverse reads for each nucleosome.

• delta a vector of numeric of length k, the updated distance between the maxima of the forward and reverse reads position densities for each nucleosome.

• w a vector of positive numerical of length k, the updated weight for each nucleosome.

• df a vector of positive numerical of length k, the updated degrees of freedom for each nucleosome.

• a a vector of positive integer of length k + 1, the updated positions, on the chromosome, delimiting the regions of the reads associated with each nucleosome.

• dim a vector of positive integer of length k, the updated number of reads associated to each nucleosome.

• rho a numeric, the acceptance probability.

Author(s)

Rawane Samb, Pascal Belleau, Astrid Deschenes

Examples

## Load dataset
data(reads_demo)

## Create a list containing the mandatory parameters
paramList <- list(kmax = 5L, nf = length(reads_demo$readsForward),
nr = length(reads_demo$readsReverse),
nbrReads = length(reads_demo$readsForward) + length(reads_demo$readsReverse),
y = sort(c(reads_demo$readsForward, reads_demo$readsReverse)),
startPSF = reads_demo$readsForward,
startPSR = reads_demo$readsReverse, lambda = 2,
zeta = 147, deltamin = 142, deltamax = 152,
minReadPos = min(c(reads_demo$readsReverse, reads_demo$readsForward)),
maxReadPos = max(c(reads_demo$readsReverse, reads_demo$readsForward)))

## Initial position of the nucleosome
muPosition <- c(73000)
muPosition

# Metropolis-Hastings move when 1 nucleosome present
result <- RJMCMC:::mhMoveK1(paramValues = paramList, kValue = 1L,
muValue = muPosition, sigmafValue = c(100), sigmarValue = c(120),
deltaValue = c(200), wValue = c(1), dfValue = c(3),
aValue = c(min(c(reads_demo$readsReverse, reads_demo$readsForward)),
max(c(reads_demo$readsReverse, reads_demo$readsForward))),
dimValue = c(length(reads_demo$readsForward) +
length(reads_demo$readsReverse)))

## Final position of the nucleosome
result$mu[1]

ArnaudDroitLab/RJMCMC documentation built on May 5, 2019, 7:06 a.m.