R/normIRS.R

In twesleyb/tidyProt: tidy statistical analysis of protein mass spectrometry

#' normIRS
#'
#' perform IRS normalization
#'
#' @export normIRS
#'
#' @importFrom data.table as.data.table
#'
#' @importFrom dplyr %>% ungroup filter group_by select
#'
#' @importFrom dplyr summarize group_split left_join 


normIRS <- function(tp, controls, robust = FALSE) {
  # perform IRS Normalization
  # Calculate average of QC samples for each experiment (ExpMean)
  tp <- dplyr::ungroup(tp)
  tp_list <- tp %>%
    dplyr::filter(Condition == controls) %>%
    dplyr::group_by(Accession, Mixture) %>%
    dplyr::summarize(
      nQC = length(Intensity),
      ExpMean = mean(Intensity), .groups="drop"
    ) %>%
    dplyr::group_split()
  # For every Protein, calculate the global mean of QC samples.
  # Calculate normalization factors to equalize GlobalMean and
  # ExpMean.
  tp_list <- lapply(tp_list, function(x) {
    x$GlobalMean <- mean(x$ExpMean, na.rm = TRUE)
    x$RobustMean <- exp(mean(log(x$ExpMean), na.rm = TRUE))
    x$NormFactor <- x$GlobalMean / x$ExpMean
    x$RobustNormFactor <- x$RobustMean / x$ExpMean
    x$NormQC <- x$NormFactor * x$ExpMean
    x$RobustNormQC <- x$RobustNormFactor * x$ExpMean
    return(x)
  })
  # Collect the data in a df
  df <- do.call(rbind, tp_list)
  # Use NormFactor's to normalize protein measurements
  tp_norm <- dplyr::left_join(tp, df, by = c("Accession", "Mixture"))
  if (robust == TRUE) {
    # Use Robust mean
    tp_norm$Intensity <- tp_norm$Intensity * tp_norm$RobustNormFactor
  } else {
    # Use Arithmetic mean
    tp_norm$Intensity <- tp_norm$Intensity * tp_norm$NormFactor
  }
  tp_norm <- tp_norm %>% dplyr::select(colnames(tp))
  tp_norm <- data.table::as.data.table(tp_norm)
  return(tp_norm)
}