karyotypeMeasures: Measures for Karyotype Heterogeneity

Description Usage Arguments Details Value Author(s) Examples

View source: R/karyotypeMeasures.R

Description

Computes measures for karyotype heterogeneity. See the Details section for how these measures are defined.

Usage

1
2
karyotypeMeasures(hmms, normalChromosomeNumbers = NULL, regions = NULL,
  exclude.regions = NULL)

Arguments

hmms

A list with aneuHMM objects or a list of files that contain such objects.

normalChromosomeNumbers

A named integer vector or matrix with physiological copy numbers, where each element (vector) or column (matrix) corresponds to a chromosome. This is useful to specify male or female samples, e.g. c('X'=2) for female samples or c('X'=1,'Y'=1) for male samples. Specify a vector if all your hmms have the same physiological copy numbers. Specify a matrix if your hmms have different physiological copy numbers (e.g. a mix of male and female samples). If not specified otherwise, '2' will be assumed for all chromosomes.

regions

A GRanges-class object containing ranges for which the karyotype measures will be computed.

exclude.regions

A GRanges-class with regions that will be excluded from the computation of the karyotype measures. This can be useful to exclude regions with artifacts.

Details

We define x as the vector of copy number states for each position. The number of HMMs is S. The measures are computed for each bin as follows:

Aneuploidy:

D = mean( abs(x-P) ), where P is the physiological number of chromosomes at that position.

Heterogeneity:

H = sum( table(x) * 0:(length(table(x))-1) ) / S

Value

A list with two data.frames, containing the karyotype measures $genomewide and $per.chromosome. If region was specified, a third list entry $regions will contain the regions with karyotype measures.

Author(s)

Aaron Taudt

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
### Example 1 ###
## Get results from a small-cell-lung-cancer
lung.folder <- system.file("extdata", "primary-lung", "hmms", package="AneuFinderData")
lung.files <- list.files(lung.folder, full.names=TRUE)
## Get results from the liver metastasis of the same patient
liver.folder <- system.file("extdata", "metastasis-liver", "hmms", package="AneuFinderData")
liver.files <- list.files(liver.folder, full.names=TRUE)
## Compare karyotype measures between the two cancers
normal.chrom.numbers <- rep(2, 23)
names(normal.chrom.numbers) <- c(1:22,'X')
lung <- karyotypeMeasures(lung.files, normalChromosomeNumbers=normal.chrom.numbers)
liver <- karyotypeMeasures(liver.files, normalChromosomeNumbers=normal.chrom.numbers)
print(lung$genomewide)
print(liver$genomewide)

### Example 2 ###
## Construct a matrix with physiological copy numbers for a mix of 5 male and 5 female samples
normal.chrom.numbers <- matrix(2, nrow=10, ncol=24,
                              dimnames=list(sample=c(paste('male', 1:5), paste('female', 6:10)),
                                            chromosome=c(1:22,'X','Y')))
normal.chrom.numbers[1:5,c('X','Y')] <- 1
normal.chrom.numbers[6:10,c('Y')] <- 0
print(normal.chrom.numbers)

### Example 3 ###
## Exclude artifact regions with high variance
consensus <- consensusSegments(c(lung.files, liver.files))
variance <- apply(consensus$copy.number, 1, var)
exclude.regions <- consensus[variance > quantile(variance, 0.999)]
## Compare karyotype measures between the two cancers
normal.chrom.numbers <- rep(2, 23)
names(normal.chrom.numbers) <- c(1:22,'X')
lung <- karyotypeMeasures(lung.files, normalChromosomeNumbers=normal.chrom.numbers,
                         exclude.regions = exclude.regions)
liver <- karyotypeMeasures(liver.files, normalChromosomeNumbers=normal.chrom.numbers,
                          exclude.regions = exclude.regions)
print(lung$genomewide)
print(liver$genomewide)

ataudt/aneufinder documentation built on Nov. 21, 2018, 10:10 a.m.