R/simplifyProteins.R

In Scavetta/ComPrAn: Complexome Profiling Analysis package

#' Simplify Raw Proteins file
#'
#'\itemize{
#' \item For rows: Keep only one row with highest Precursor Area in cases where
#' for a single Protein Group Accession in a single fraction there are multiple 
#' rows with the same combination of Sequence, Mods and Charge
#' \item For cols: remove columns that are not neccessary any more
#'}
#' @param .data a dataframe
#' @param direction character, rows, cols or both
#'
#' @return a dataframe
#' @export
#'
#' @examples
#' ##Use example peptide data set, read in and clean data
#' inputFile <- system.file("extData", "data.txt", package = "ComPrAn")
#' peptides <- peptideImport(inputFile)
#' peptides <- cleanData(peptides, fCol = "Search ID")
#' ## separate chemical modifications and labelling into separate columns
#' peptides <- splitModLab(peptides) 
#' ## remove unneccessary columns, simplify rows
#' peptides <- simplifyProteins(peptides) 
#'
simplifyProteins <- function(.data, direction = c("rows", "cols")) {
    colsToSelect <- c("Fraction", "Precursor Area", "Protein Group Accessions",
                        "Protein Descriptions", "Mods", "Labels",
                        "Charge", "Sequence",
                        "isLabel", "UniqueCombinedID_A","UniqueCombinedID_B")
    
    if(!all(colsToSelect %in% names(.data))) {
        stop('Not all columns found, you must run splitModLab function first.') 
    }
    
    if ("cols" %in% direction) {
        # Remove unnecessary columns to save memory
        .data %>%
            dplyr::select(colsToSelect) -> .data
    }
    
    if ("rows" %in% direction) {
        .data %>%
            dplyr::group_by(`Protein Group Accessions`, UniqueCombinedID_B,
                            Fraction) %>%
            dplyr::filter(`Precursor Area` == max(`Precursor Area`),
                            !duplicated(`Precursor Area`)) -> .data
    }
}