R/parse_BN_crosstabFile.R

In CogDisResLab/KRSA: KRSA: Kinome Random Sampling Analyzer

#' Parse bionavigator crosstab files
#'
#' Main function that parses bionavigator crosstab view files. Takes in path
#' to file and type (either Signal or SignalSaturation)
#'
#' @param file_path path to BN crosstab file
#' @param type Name of the value (either Signal or SignalSaturation)
#'
#' @return tbl_df
#'
#' @export
#'
#' @examples
#' TRUE
parse_BN_crosstabFile <- function(file_path, type = c(
                                      "Signal",
                                      "SignalSaturation"
                                  )) {
    # Parse a BN crosstab file
    # Created 2020-02-14
    # Last Updated 2020-02-14

    df <- utils::read.csv(file_path,
        sep = "\t", header = F, stringsAsFactors = F
    )

    meta_rows <- which(df[, 1] == "")
    epmty_col <- which(df[meta_rows[1], ] == "")

    meta_info <- df[meta_rows, (epmty_col[length(epmty_col)] + 1):ncol(df)]

    rows_end <- nrow(df)
    rows_start <- meta_rows[(length(meta_rows))] + 2

    data <- df[rows_start:rows_end, ]
    data <- data[, c(1, (epmty_col[length(epmty_col)] + 2):ncol(data))]

    df2 <- data.frame(t(data[-1]), stringsAsFactors = F)
    colnames(df2) <- data[, 1]
    colnames(df2) <- make.unique(colnames(df2))

    df3 <- data.frame(t(meta_info[-1]), stringsAsFactors = F)
    colnames(df3) <- meta_info[, 1]
    cbind(df3, df2) -> df4

    df4 %>%
        tidyr::pivot_longer(cols = (length(meta_rows) + 1):ncol(.), names_to = "Peptide", values_to = type) -> tidydata

    tidydata$`Exposure time` <- as.numeric(tidydata$`Exposure time`)
    tidydata$Cycle <- as.numeric(tidydata$Cycle)

    if (type == "Signal") {
        tidydata$Signal <- as.numeric(tidydata$Signal)
    } else {
        tidydata$SignalSaturation <- as.numeric(tidydata$SignalSaturation)
    }

    tidydata %>%
        dplyr::rename(
            SampleName = `Sample name`,
            ExposureTime = `Exposure time`
        ) -> tidydata

    return(tidydata)
}