R/write_matrix_proteins.R

In SWATH2stats: Transform and Filter SWATH Data for Statistical Packages

#' Writes out an overview matrix of summed signals per protein identifier
#' (lines) over run_id(columns).
#'
#' Writes out an overview matrix on protein level of a supplied (unfiltered or
#' filtered) OpenSWATH results data frame. The protein quantification is achieved
#' by summing the areas under all 6 transitions per precursor, summing all
#' precursors per FullPeptideName and all FullPeptideName signals per
#' ProteinName entry.
#' This function does not select consistently quantified or top peptides but
#' sums all signals availabe that may or may not originate from the same set of
#' peptides across different runs. A more detailed overview can be generated
#' using the function write_matrix_peptides().
#' Peptide selection can be achieved upstream using e.g. the functions
#' filter_mscore_requant(), filter_on_max_peptides() and
#' filter_on_min_peptides().
#'
#' @param data A data frame containing annotated OpenSWATH/pyProphet data.
#' @param write_csv  Option to determine if table should be written 
#'   automatically into csv file.
#' @param fun_aggregate  What function to use when aggregating the set of 
#'   intensities (sum or mean)?. Default: sum.
#' @param filename  File base name of the .csv matrix written out to the working
#'   folder
#' @param rm_decoy  Logical whether decoys will be removed from the data
#'   matrix. Defaults to FALSE. It's sometimes useful to know how decoys behave
#'   across a dataset and how many you allow into your final table with the
#'   current filtering strategy.
#' @return the peptides as a matrix, also output .csv matrix is written to the
#'   working folder
#' @author Moritz Heusel
#' @examples{
#'  data("OpenSWATH_data", package="SWATH2stats")
#'  data("Study_design", package="SWATH2stats")
#'  data <- sample_annotation(OpenSWATH_data, Study_design)
#'  written <- write_matrix_proteins(data)
#'  }
#' @importFrom reshape2 dcast
#' @importFrom stats aggregate
#' @export
write_matrix_proteins <- function(data, 
                                  write_csv = FALSE, 
                                  fun_aggregate = "sum",
                                  filename = "SWATH2stats_overview_matrix_proteinlevel.csv",
                                  rm_decoy = FALSE) {
    if (rm_decoy == TRUE) {
        data <- subset(data, data$decoy == 0)
    }
    if(fun_aggregate == "sum"){
        data.protein.sum.all <- aggregate(data[, "Intensity"], 
                                         by = list(data$ProteinName, data$run_id), sum)
        colnames(data.protein.sum.all) <- c("ProteinName", "run_id", "Intensity.all.sum")
        data.protein.aggr.table <- dcast(data.protein.sum.all, ProteinName ~ run_id, 
                                        value.var = "Intensity.all.sum", 
                                        fun.aggregate = sum)
    }
    if(fun_aggregate == "mean"){
        data.protein.mean.all <- aggregate(data[, "Intensity"], 
                                         by = list(data$ProteinName, data$run_id), mean)
        colnames(data.protein.mean.all) <- c("ProteinName", "run_id", "Intensity.all.mean")
        data.protein.aggr.table <- dcast(data.protein.mean.all, ProteinName ~ run_id, 
                                        value.var = "Intensity.all.mean", 
                                        fun.aggregate = mean)
        
    }
      if (write_csv) {
        write.csv(data.protein.aggr.table, file = filename, 
                  row.names = FALSE, quote = FALSE)
        message("Protein overview matrix ", filename, " written to working folder.",
            "\n")
    }
    return(data.protein.aggr.table)
}