rjmcmc: Nucleosome positioning mapping on a segment

In RJMCMCNucleosomes: Bayesian hierarchical model for genome-wide nucleosome positioning with high-throughput short-read data (MNase-Seq)

Description

Use of a fully Bayesian hierarchical model for chromosome-wide profiling of nucleosome positions based on high-throughput short-read data (MNase-Seq data). Beware that for a genome-wide profiling, each chromosome must be treated separatly. This function is optimized to run on segments that are smaller sections of the chromosome.

Usage

rjmcmc(reads, seqName = NULL, nbrIterations, kMax, lambda = 3, minInterval,
  maxInterval, minReads = 5, adaptIterationsToReads = TRUE, vSeed = -1,
  saveAsRDS = FALSE)

Arguments

reads	a GRanges containing forward and reverse reads. Beware that the start position of a reverse read is always higher that the end positition.
seqName	a character string containing the label of the chromosome, present in the GRanges object, that will be used. The NULL value is accepted when only one seqname is present in the GRanges; the only seqname present will be used. Default: NULL.
nbrIterations	a positive integer or numeric, the number of iterations. Non-integer values of nbrIterations will be casted to integer and truncated towards zero.
kMax	a positive integer or numeric, the maximum number of degrees of freedom per region. Non-integer values of kMax will be casted to integer and truncated towards zero.
lambda	a positive numeric, the theorical mean of the Poisson distribution. Default: 3.
minInterval	a numeric, the minimum distance between two nucleosomes.
maxInterval	a numeric, the maximum distance between two nucleosomes.
minReads	a positive integer or numeric, the minimum number of reads in a potential canditate region. Non-integer values of minReads will be casted to integer and truncated towards zero. Default: 5.
adaptIterationsToReads	a logical indicating if the number of iterations must be modified in function of the number of reads. Default: TRUE.
vSeed	a integer. A seed used when reproducible results are needed. When a value inferior or equal to zero is given, a random integer is used. Default: -1.
saveAsRDS	a logical. When TRUE, a RDS file containing the complete output of the c++ rjmcmc() function is created. Default : FALSE.

Value

a list of class "rjmcmcNucleosomes" containing:

• call the matched call.

• k a integer, the final estimation of the number of nucleosomes. 0 when no nucleosome is detected.

• mu a GRanges containing the positions of the nucleosomes and '*' as strand. The seqnames of the GRanges correspond to the seqName input value. NA when no nucleosome is detected.

• k_max a integer, the maximum number of nucleosomes obtained during the iteration process. NA when no nucleosome is detected.

Author(s)

Rawane Samb, Pascal Belleau, Astrid Deschenes

Examples

## Loading dataset
data(reads_demo_01)

## Nucleosome positioning, running both merge and split functions
result <- rjmcmc(reads = reads_demo_01, seqName = "chr_SYNTHETIC",
            nbrIterations = 1000, lambda = 2, kMax = 30,
            minInterval = 146, maxInterval = 292, minReads = 5,
            vSeed = 10113, saveAsRDS = FALSE)

## Print the final estimation of the number of nucleosomes
result$k

## Print the position of nucleosomes
result$mu

## Print the maximum number of nucleosomes obtained during the iteration
## process
result$k_max

RJMCMCNucleosomes documentation built on Nov. 8, 2020, 8:20 p.m.