Genome segmentation based on feature density — segmentDensity • nullranges

This function allows for various methods (see type) of segmenting based on the density of features x.

segmentDensity(x, n, L_s = 1e+06, exclude = NULL, type = c("cbs", "hmm"))

Arguments

x: the input GRanges, e.g. genes
n: the number of states
L_s: segment length
exclude: GRanges of excluded region
type: the type of segmentation, either Circular Binary Segmentation "cbs" (which will use 'DNAcopy' to segment) or Hidden Markov Model "hmm" (which will use 'RcppHMM'). These packages are not imported by nullranges, but must be installed by the user

Value

a GRanges with metadata columns containing:

state segmentation state
counts average number of genes

References

Circular binary segmentation (CBS):

Olshen, A. B., E. S. Venkatraman, R. Lucito, and M. Wigler. 2004. "Circular binary segmentation for the analysis of array-based DNA copy number data." Biostatistics 5 (4): 557–72.

Hidden Markov Model from RcppHMM:

Roberto A. Cardenas-Ovando, Julieta Noguez, and Claudia Rangel-Escareno. "Rcpp Hidden Markov Model." CRAN R package.

Examples


n <- 10000
library(GenomicRanges)
gr <- GRanges("chr1", IRanges(round(
  c(runif(n/4,1,991), runif(n/4,1001,3991),
    runif(n/4,4001,4991), runif(n/4,7001,9991))),
  width=10), seqlengths=c(chr1=10000))
gr <- sort(gr)
exclude <- GRanges("chr1", IRanges(5001,6000), seqlengths=c(chr1=10000))
seg <- segmentDensity(gr, n=3, L_s=100, exclude=exclude, type="cbs")
#> Analyzing: Sample.1