R/population_annotate_functions.R

In kdaily/MEMA: Functions to preprocess and normalize MEMA data

#Functions to support MEMAs printed in 8 well plates


#'Melt raw population dataset to combine wells
#'
#'\code{melt8Well} Converts a Tecan ArrayPro data file to a data.table with properly labeled columns.
#'
#'@param DT a data.table with columns of population intensity values in columns named by barcode, well, wavelength and data type(Raw/Net/Background)
#'@param swap a logical on whether the data for the A and B rows should be swapped. This must be yes for 8 well data coming from the Array Pro software.
#'@return A data.table in long format with columns of intensity readings for #' each data type and wavelength, barcode and well.
#'@section Usage: This function reformats Array Pro results files that
#' come from analyzing images from multiple channels, wells and plates.
#' The column names are encoded as follows:
#' Datatype is at the beginning of the column name as Raw, Net or
#' Background. Barcode follows the curly brace symbol. Wavelength follows
#'  the - symbol after the barcode. Well follows the - symbol after the
#'  wavelength. The A and B well labels are swapped to align the Tecan
#'  labeling to the actual well names.
#'  @importFrom "data.table" ":="

#' @export
melt8Well<-function(DT,swap = TRUE){

  intNames<-grep("Background|Net|Raw",colnames(DT), value = TRUE)
  intNameType<-lapply(intNames,function(intName){
    return(strsplit(intName,split=" ")[[1]][1])
  })
  intNameWL<-lapply(intNames,function(intName){
    return(strsplit(intName,split="-")[[1]][2])
  })
  intNameWell<-lapply(intNames,function(intName){
    return(strsplit(intName,split="-")[[1]][3])
  })
  intNameBarcode<-lapply(intNames,function(intName){
    tmp<-strsplit(intName,split="[{]")[[1]][2]
    return(strsplit(tmp,split="-")[[1]][1])
  })
  newIntNames<-make.names(paste(intNameType,intNameWL,intNameWell,intNameBarcode))
  data.table::setnames(DT,colnames(DT[,intNames, with=FALSE]),newIntNames)
  meltDT<-reshape2::melt(DT,measure=newIntNames,variable="TypeWLWellBarcode",value="Intensity", variable.factor=FALSE)
  meltDT$Type<-unlist(lapply(meltDT$TypeWLWellBarcode,function(x){
    return(strsplit(x,split="[.]")[[1]][1])
  }))
  meltDT$WL<-unlist(lapply(meltDT$TypeWLWellBarcode,function(x){
    return(strsplit(x,split="[.]")[[1]][2])
  }))
  meltDT$Well<-unlist(lapply(meltDT$TypeWLWellBarcode,function(x){
    well<-strsplit(x,split="[.]")[[1]][3]
    wellChars<-strsplit(well,"")
    Well<-paste0(wellChars[[1]][1],"0",wellChars[[1]][2])
    return(Well)
  }))
  meltDT$Barcode<-unlist(lapply(meltDT$TypeWLWellBarcode,function(x){
    return(strsplit(x,split="[.]")[[1]][4])
  }))
  meltDT[,TypeWLWellBarcode:=NULL]
  meltDT$Type<-paste(meltDT$Type,meltDT$WL,sep=".")
  #Delete the Wavelength only column
  meltDT[,WL:=NULL]
  data.table::setnames(meltDT,colnames(meltDT),make.names(colnames(meltDT)))

  #wide format creating columns of data organized by type and wavelength
  DT<-reshape2::dcast(meltDT,...~Type,value.var="Intensity")
  if (swap){
    #Swap A and B well names to align Tecan labeling to the actual well names
    DT$Well <- gsub("A","T",DT$Well)
    DT$Well <- gsub("B","A",DT$Well)
    DT$Well <- gsub("T","B",DT$Well)
  }

  DT<-data.table::data.table(DT,key=c("Well","Barcode"))
  return(DT)
}