R/countMatrix.R

In maftools: Summarize, Analyze and Visualize MAF Files

#' Generates count matrix of mutations.
#'
#' @description Generates a count matrix of mutations. i.e, number of mutations per gene per sample.
#' @param maf an MAF object generated by \code{\link{read.maf}}
#' @param includeSyn whether to include sysnonymous variants in ouput matrix. Default FALSE
#' @param countOnly Default NULL - counts all variants. You can specify type of 'Variant_Classification' to count. For e.g, countOnly = 'Splice_Site' will generates matrix for only Splice_Site variants.
#' @param removeNonMutated Logical Default \code{TRUE}, removes samples with no mutations from the matrix.
#' @return Integer Matrix
#' @seealso \code{\link{getFields}} \code{\link{getGeneSummary}} \code{\link{getSampleSummary}}
#' @examples
#' laml.maf <- system.file("extdata", "tcga_laml.maf.gz", package = "maftools")
#' laml <- read.maf(maf = laml.maf)
#' ##Generate matrix
#' mutCountMatrix(maf = laml)
#' ##Generate count matrix of Splice_Site mutations
#' mutCountMatrix(maf = laml, countOnly = 'Splice_Site')
#' @export

mutCountMatrix = function(maf, includeSyn = FALSE, countOnly = NULL, removeNonMutated = TRUE){

  #Synonymous variants
  maf.silent = maf@maf.silent
  #Main data
  maf.dat = maf@data

  if(includeSyn){
    maf.dat = rbind(maf.dat, maf.silent, fill = TRUE)
  }

  if(!is.null(countOnly)){
    maf.dat = maf.dat[Variant_Classification %in% countOnly]
    if(nrow(maf.dat) == 0){
      stop(paste0("No mutations available for ", countOnly, collapse = ''))
    }
  }

  maf.dat = maf.dat[,.N,.(Hugo_Symbol, Tumor_Sample_Barcode)]

  if(!removeNonMutated){
    maf.dat.mat = data.table::dcast(data = maf.dat, Hugo_Symbol ~ Tumor_Sample_Barcode, fill = 0, value.var = 'N', drop = FALSE)
  }else{
    maf.dat.mat = data.table::dcast(data = maf.dat, Hugo_Symbol ~ Tumor_Sample_Barcode, fill = 0, value.var = 'N')
  }


  maf.dat.mat$tot = rowSums(maf.dat.mat[,2:ncol(maf.dat.mat)])
  maf.dat.mat = maf.dat.mat[order(tot, decreasing = TRUE)]
  maf.dat.mat[,tot := NULL]
  data.table::setDF(x = maf.dat.mat)
  rownames(maf.dat.mat) = maf.dat.mat$Hugo_Symbol
  maf.dat.mat = maf.dat.mat[,-1]

  maf.temp.copy = maf.dat.mat #temp copy of original unsorted numeric coded matrix

  maf.dat.mat[maf.dat.mat != 0] = 1 #replacing all non-zero integers with 1 improves sorting (& grouping)
  tmaf = t(maf.dat.mat) #transposematrix
  maf.dat.mat = t(tmaf[do.call(order, c(as.list(as.data.frame(tmaf)), decreasing = TRUE)), ]) #sort

  maf.temp.copy = maf.temp.copy[rownames(maf.dat.mat),] #organise original matrix into sorted matrix
  maf.temp.copy = maf.temp.copy[,colnames(maf.dat.mat)]
  maf.dat.mat = maf.temp.copy

  return(maf.dat.mat)
}