R/HMDB_metabolites.R

In sipss/NIHSlcms: Signal processing of LC-MS metabolomics

#' HMDB metabolites in positive ionization
#'
#' `assignation_pos_HMDB` is a function to identify the features in negative
#' ionization. It assumes protonation of molecules (M - H) from the feature
#' table and compares with mass of metabolites contained in the The Human
#' Metabolome Database [HMDB_db].
#'
#' @param feature_table A data frame containing feature in rows and samples in
#'   columns. A column called "mz" is required with the corresponding mass of
#'   each feature.
#'
#' @return The same feature table with two extra column with matched HMDB codes
#'   and metabolites. More than one assignation is allowed.
#' @export
#' @usage
#' assignation_pos_HMDB(feature_table)
#' @examples
#' \dontrun{
#' FT <- assignation_pos_HMDB(feature_table)
#' head(FT)
#' }

assignation_pos_HMDB <- function(feature_table){

  HMDB_db <- NULL
  utils::data("HMDB_db", package = "AlpsLCMS", envir = environment())

  output_assignation_list <- HMDB_db[NULL,]
  MZs <- rownames(feature_table)
  FT <- feature_table
  FT$HMDB <- NA
  FT$Assignation <- NA

  for (i in 1: nrow(feature_table)) {
    k = feature_table[i, "mz"]
    lower_mz_edge <- k - 0.002
    higher_mz_edge <- k + 0.002

    ind <- intersect(which(HMDB_db$M.plus.H < higher_mz_edge),
                     which(HMDB_db$M.plus.H > lower_mz_edge))
    # assignation_list <- as.data.frame(HMDB_db[ind,])
    if (length(ind) == 0){

      FT[i,"HMDB"] <- NA
      FT[i,"Assignation"] <- NA


    } else {

      FT[i,"HMDB"] <- paste(as.character(HMDB_db[ind, "ENTRY"]), collapse="--- ")
      FT[i,"Assignation"] <- paste(as.character(HMDB_db[ind, "NAME"]), collapse="---")
    }

    output_assignation_list <- rbind(output_assignation_list, HMDB_db[ind,])

  }

  return(FT)
}


#' HMDB metabolites in negative ionization
#'
#' `assignation_neg_HMDB` is a function to identify the features in negative
#' ionization. It assumes deprotonation of molecules (M - H) from the feature
#' table and compares with mass of metabolites contained in the The Human
#' Metabolome Database [HMDB_db].
#'
#' @param feature_table A data frame containing feature in rows and samples in
#'   columns. First and second columns need to be the feature ID and mass.
#'
#' @return The same feature table with two extra column with matched HMDB codes
#'   and metabolites. More than one assignation is allowed.
#' @export
#' @examples
#' \dontrun{
#' FT <- assignation_neg_HMDB(feature_table)
#' head(FT)
#' }

assignation_neg_HMDB <- function(feature_table){

  HMDB_db <- NULL
  utils::data("HMDB_db", package = "AlpsLCMS", envir = environment())

  output_assignation_list <- HMDB_db[NULL,]
  MZs <- rownames(feature_table)
  FT <- feature_table
  FT$HMDB <- NA
  FT$Assignation <- NA

  for (i in 1: nrow(feature_table)) {
    k = feature_table[i, "mz"]
    lower_mz_edge <- k - 0.002
    higher_mz_edge <- k + 0.002

    ind <- intersect(which(HMDB_db$M.minus.H < higher_mz_edge),
                     which(HMDB_db$M.minus.H > lower_mz_edge))
    # assignation_list <- as.data.frame(HMDB_db[ind,])
    if (length(ind) == 0){

      FT[i,"HMDB"] <- NA
      FT[i,"Assignation"] <- NA


    } else {

      FT[i,"HMDB"] <- paste(as.character(HMDB_db[ind, "ENTRY"]), collapse="--- ")
      FT[i,"Assignation"] <- paste(as.character(HMDB_db[ind, "NAME"]), collapse="---")
    }

    output_assignation_list <- rbind(output_assignation_list, HMDB_db[ind,])

  }

  return(FT)
}

#' The Human Metabolome DataBase template: Table with metabolites along with
#' their protonated/deprotonated masses
#'
#' @name HMDB_db
#' @docType data
#' @references \url{hmdb.ca}
#' @keywords data
NULL