maftools (version 0.99.30)

inferHeterogeneity: Clusters variants based on Variant Allele Frequencies (VAF).

Description

takes output generated by read.maf and clusters variants to infer tumor heterogeneity. This function requires VAF for clustering and density estimation. VAF can be on the scale 0-1 or 0-100. Optionally if copy number information is available, it can be provided as a segmented file (e.g, from Circular Binary Segmentation). Those variants in copy number altered regions will be ignored.

Usage

inferHeterogeneity(maf, tsb = NULL, top = 5, vafCol = NULL,
  dirichlet = FALSE, segFile = NULL, ignChr = NULL, minVaf = 0,
  maxVaf = 1)

Arguments

maf

an MAF object generated by read.maf

tsb

specify sample names (Tumor_Sample_Barcodes) for which clustering has to be done.

top

if tsb is NULL, uses top n number of most mutated samples. Defaults to 5.

vafCol

manually specify column name for vafs. Default looks for column 't_vaf'

dirichlet

If TRUE uses nonparametric dirichlet process for clustering. Default FALSE, uses finite mixture models.

segFile

path to CBS segmented copy number file. Column names should be Sample, Chromosome, Start, End, Num_Probes and Segment_Mean (log2 scale).

ignChr

ignore these chromosomes from analysis. e.g, sex chromsomes chrX, chrY. Default NULL.

minVaf

filter low frequency variants. Low vaf variants maybe due to sequencing error. Default 0. (on the scale of 0 to 1)

maxVaf

filter high frequency variants. High vaf variants maybe due to copy number alterations or impure tumor. Default 1. (on the scale of 0 to 1)

Value

list of clustering tables.

Details

This function clusters variants based on VAF to estimate univariate density and cluster classification. There are two methods available for clustering. Default using parametric finite mixture models and another method using nonparametric inifinite mixture models (Dirichlet process).

References

Chris Fraley and Adrian E. Raftery (2002) Model-based Clustering, Discriminant Analysis and Density Estimation Journal of the American Statistical Association 97:611-631

Jara A, Hanson TE, Quintana FA, Muller P, Rosner GL. DPpackage: Bayesian Semi- and Nonparametric Modeling in R. Journal of statistical software. 2011;40(5):1-30.

Olshen AB, Venkatraman ES, Lucito R, Wigler M. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics. 2004;5(4):557-72.

See Also

plotClusters

Examples

# NOT RUN {
laml.maf <- system.file("extdata", "tcga_laml.maf.gz", package = "maftools")
laml <- read.maf(maf = laml.maf, removeSilent = TRUE, useAll = FALSE)
TCGA.AB.2972.clust <- inferHeterogeneity(maf = laml, tsb = 'TCGA.AB.2972', vafCol = 'i_TumorVAF_WU')

# }