R/clean_up.R

In ricoderks/lipidomics: Lipidomics workflow

#' @title Clean up tibble from MS-DIAL
#'
#' @description Clean up the columns and column names of the tibble after
#'     reading the MS-DIAL result files.
#'
#' @param lipid_data The tibble.
#'
#' @return Returns a tibble
#'
#' @importFrom dplyr rename mutate select
#' @importFrom tidyselect matches
#' @importFrom tidyr unnest
#' @importFrom rlang .data
#'
#' @author Rico Derks
#'
clean_up <- function(lipid_data) {
  # make a single dataframe
  lipid_data <- lipid_data %>%
    dplyr::select(.data$polarity, .data$raw_data) %>%
    tidyr::unnest(c(.data$polarity, .data$raw_data))

  # remove spaces at the front or at the back of the column name
  colnames(lipid_data) <- trimws(colnames(lipid_data))

  column_names <- colnames(lipid_data)

  if("Simple dot product" %in% column_names) {
    # version > 5.1
    # rename some columns in the data frame for ease of access later on.
    lipid_data <- lipid_data %>%
      dplyr::rename(AlignmentID = .data$`Alignment ID`,
                    AverageRT = .data$`Average Rt(min)`,
                    AverageMZ = .data$`Average Mz`,
                    ion = .data$`Adduct type`,
                    LipidName = .data$`Metabolite name`,
                    LipidClass = .data$Ontology,
                    DotProduct = .data$`Simple dot product`,
                    RevDotProduct = .data$`Reverse dot product`,
                    TotalScore = .data$`Total score`,
                    FragPresence = .data$`Matched peaks percentage`,
                    RefFile = .data$`Spectrum reference file name`,
                    MSMSspectrum = .data$`MS/MS spectrum`) %>%
      dplyr::mutate(DotProduct = .data$DotProduct * 100,
                    RevDotProduct = .data$RevDotProduct * 100,
                    scale_DotProduct = .data$DotProduct / 10,
                    scale_RevDotProduct = .data$RevDotProduct / 10,
                    my_id = paste(.data$polarity, "_", .data$AlignmentID, sep = ""),
                    ion = factor(.data$ion),
                    polarity = factor(.data$polarity)) %>%
      dplyr::select(.data$my_id, .data$AlignmentID, .data$AverageRT, .data$AverageMZ, .data$ion, .data$LipidName, .data$LipidClass,
                    .data$DotProduct, .data$scale_DotProduct, .data$RevDotProduct, .data$scale_RevDotProduct,
                    .data$FragPresence, .data$TotalScore, .data$polarity, .data$MSMSspectrum,
                    matches("^([qQ][cC]pool|[sS]ample|[bB]lank)_.*[0-9]{3}$"))
  } else {
    # version 4.x
    # rename some columns in the data frame for ease of access later on.
    lipid_data <- lipid_data %>%
      dplyr::rename(AlignmentID = .data$`Alignment ID`,
                    AverageRT = .data$`Average Rt(min)`,
                    AverageMZ = .data$`Average Mz`,
                    ion = .data$`Adduct type`,
                    LipidName = .data$`Metabolite name`,
                    LipidClass = .data$Ontology,
                    DotProduct = .data$`Dot product`,
                    RevDotProduct = .data$`Reverse dot product`,
                    TotalScore = .data$`Total score`,
                    FragPresence = .data$`Fragment presence %`,
                    RefFile = .data$`Spectrum reference file name`,
                    MSMSspectrum = .data$`MS/MS spectrum`) %>%
      dplyr::mutate(scale_DotProduct = .data$DotProduct / 10,
                    scale_RevDotProduct = .data$RevDotProduct / 10,
                    my_id = paste(.data$polarity, "_", .data$AlignmentID, sep = ""),
                    ion = factor(.data$ion),
                    polarity = factor(.data$polarity)) %>%
      dplyr::select(.data$my_id, .data$AlignmentID, .data$AverageRT, .data$AverageMZ, .data$ion, .data$LipidName, .data$LipidClass,
                    .data$DotProduct, .data$scale_DotProduct, .data$RevDotProduct, .data$scale_RevDotProduct,
                    .data$FragPresence, .data$TotalScore, .data$polarity, .data$MSMSspectrum,
                    matches("^([qQ][cC]pool|[sS]ample|[bB]lank)_.*[0-9]{3}$"))
  }
  return(lipid_data)
}