R/CCAverage.R

In WHG1990/CCTools: A suite of tools for analysing CC-seq data.

#' @title CCAverage
#' @description Averages multiple Fullmaps together, by averaging HpM.
#' @param path.name A character string defining folder containing individual FullMaps.
#' @param exp.name Manually specify a name for the outputted pool file. Defaults to a concatenation of individual file names.
#' @param out.mode output in "txt", "rds", or "both" mode.
#' @examples
#' @author Will Gittens
#' @import data.table
#' @export
CCAverage <- function(path.name, exp.name, out.mode = "txt") {
  library("stringr")
  library("data.table")
  setwd(path.name)
  DSBList.2<- list()
  file.list <- list.files(pattern = ".rds|.txt")
  abbrev.list <- str_match(file.list, "FullMap.(.*?).txt")[,2]
  temp <- str_match(file.list, "(.*?).rds")[,2]
  abbrev.list[is.na(abbrev.list)] <- temp[!is.na(temp)]
  ext.list <- substr(file.list, nchar(file.list)-2, nchar(file.list))
  ################# data loading
  for (g in 1:length(file.list)) {
    if(ext.list[g] == "txt"){
      DSBList <- fread(file.list[g], sep = "\t", header=TRUE, col.names = c("Chr", "Pos", "Watson", "Crick"))
      DSBList.2[[g]] <- DSBList
      rm(DSBList)
    }
    if(ext.list[g] == "rds"){
      load(file.list[g])
      if(exists("DSBList.bf")){DSBList <- DSBList.bf; rm(DSBList.bf)}
      if(exists("DSBList.pooled")){DSBList <- DSBList.pooled; rm(DSBList.pooled)}
      if(exists("pooled.DSBList")){DSBList <- pooled.DSBList; rm(pooled.DSBList)}
      if(exists("dup.DSBList")){DSBList <- dup.DSBList; rm(dup.DSBList)}
      if(exists("undup.DSBList")){DSBList <- undup.DSBList; rm(undup.DSBList)}
      if("list" %in% class(DSBList)) {stop("ERROR: one or more RDS files contains more than one Fullmap. CCPool only works on inidividual Fullmaps.")}
      DSBList.2[[g]] <- DSBList
      rm(DSBList)
    }
  }
  DSBList <- DSBList.2; rm(DSBList.2)

  Mreads <- MreadR(DSBList)
  DSBList <- mapply(function(x, y){
    x$Watson <- x$Watson/y
    x$Crick <- x$Crick/y
    return(x)
  }, DSBList, Mreads, SIMPLIFY = F)

  DSBList <- lapply(DSBList, data.table)
  DSBList <- lapply(DSBList, setkey, Chr, Pos)
  DSBList.pooled <- DSBList[1]
  for (i in 2:length(DSBList)) {
    DSBList.pooled[[1]] <- merge(DSBList.pooled[[1]], DSBList[[i]], by = c("Chr", "Pos"), all = T)
    DSBList.pooled <- lapply(DSBList.pooled, function(x){
      x[is.na(x)] <- 0L
      return(x)
    })

      DSBList.pooled <- lapply(DSBList.pooled, function(x){
      x$Watson <- x$Watson.x + x$Watson.y
      x$Crick <- x$Crick.x  + x$Crick.y
      return(x)
    })
    DSBList.pooled <- lapply(DSBList.pooled, function(x){
      x$Watson.x <- NULL;  x$Watson.y <- NULL;  x$Crick.x <- NULL;  x$Crick.y <- NULL
      return(x)
    })
  }
  DSBListNames <- paste(abbrev.list, collapse = "_")
  DSBListNames <- paste0(DSBListNames, "_Average")
  DSBList <- DSBList.pooled
  nfiles <- length(DSBList)
  Mreads <- MreadR(DSBList)
  DSBList <- mapply(function(x, y){
    x$Watson <- round(x$Watson/y,4)
    x$Crick <- round(x$Crick/y,4)
    return(x)
  }, DSBList, Mreads, SIMPLIFY = F)
  if(!missing(exp.name)){DSBListNames <- exp.name}
  ifelse(!dir.exists(file.path(path.name, "Average")), dir.create(file.path(path.name, "Average")), FALSE); setwd(file.path(path.name, "Average"))
  if(out.mode == "rds"){save(DSBList, DSBListNames, nfiles, MMreads, Mreads, file = paste0(DSBListNames[1], ".rds"))}
  if(out.mode == "fullmap" | out.mode == "txt"){
    CCWrite(DSBList. = DSBList, filenames = DSBListNames)
  }
  if(out.mode == "both"){
    save(DSBList, DSBListNames, nfiles, MMreads, Mreads, file = paste0(DSBListNames[1], ".rds"))
    CCWrite(DSBList. = DSBList, filenames = DSBListNames)
  }
}