R/filterClusters.R

In ibludau/clusterMapper:

#' Mapped cluster filter
#'
#' @description Filter mapped clusters by following criteria:
#' \itemize{
#'  \item Cluster must be main cluster of the transition group.
#'  \item Main cluster frequency must be >= min_main_cluster_freq.
#'  \item Main cluster should not be zero (non-defined).
#' }
#' @param table data.table with original cluster information
#' @param min_main_cluster_freq Numeric between 0 and 1. Minimum frequency of a
#' transition group id matched to the same cluster. Default is 0.6.
#' @return data.table Same as input table but including extra columns about
#' run coverage information and filtered by \code{min_main_cluster_freq}.
#' @import data.table
#' @export
#'
#' @examples
#' data <- fullDataMapped
#' filteredData <- filterData(data,
#'                            min_main_cluster_freq = 0.6)
filterClusters <- function(table, min_main_cluster_freq = 0.6){
  if (! all(class(table) == c("data.table","data.frame"))) {
    stop("The provided input table is not a data.table. Please provide an output table from mapClusters.")
  }
  if (nrow(table) == 0) {
    stop("The provided input table is empty.")
  }
  if (! all(c("transition_group_id","run_id","RT","Collapsed_Peptide","Cluster","new_cluster") %in% names(table))) {
    stop("The provided input table does not contain all required columns. Please provide an output table from mapClusters.")
  }
  if ((min_main_cluster_freq > 1) | (min_main_cluster_freq < 0)) {
    stop("The provided min_main_cluster_freq value is not valid. Please choose a value between 0 and 1.")
  }

  features <- copy(table)
  # number of runs per transition group
  features[, run_count_per_tg := uniqueN(run_id), by = c("transition_group_id")]
  # main cluster per transition group
  getMostFrequentCluster <- function(X){as.numeric(names(sort(table(X),decreasing=TRUE))[1])}
  features[ , `:=` (main_cluster_of_tg = getMostFrequentCluster(new_cluster)) , by = c("Collapsed_Peptide","transition_group_id") ]
  # determine main cluster prevalance
  featuresFiltered <- subset(features, new_cluster == main_cluster_of_tg)
  # determine frequency of unique cluster assignment per transition group
  featuresFiltered[, run_count_per_tg_main_cluster := uniqueN(run_id), by = c("transition_group_id","new_cluster")]
  featuresFiltered[, main_cluster_tg_freq := run_count_per_tg_main_cluster/run_count_per_tg]
  featuresFiltered <- subset(featuresFiltered, main_cluster_tg_freq >= min_main_cluster_freq)
  # remove unassigned clusters
  featuresFiltered <- subset(featuresFiltered, new_cluster != 0)
  return(featuresFiltered)
}