In bihealth/metaquac: Metabolomics Targeted Quality Control

Import

Import of r params$kit data of measurement type r params$measurement_type.

Data files to import:

# Data files
pander::pander(params$data_files)

# Measurement Type
meas_type <- params$measurement_type

# Import batches
num_batches <- length(params$data_files)
imports <- vector(mode = "list", length = num_batches)
conc_col_name <- vector(mode ="character", length = num_batches)
for (i in seq_len(num_batches)){
  cat(paste0("\n\nImporting batch `", names(params$data_files)[i], "`... "))
  import <- import_biocrates_unit(data_files = params$data_files[[i]],
                                  zero2na = params$zero2na,
                                  measurement_type = meas_type,
                                  pool_indicator = POOL_INDICATOR)

  # Add batch variable
  import$Batch <- names(params$data_files)[i]

  # Collect concentration column names to check unit consistency
  conc_col_name[i] <- grep(TABLE_TYPES[1], names(import), value = TRUE)
  names(conc_col_name)[i] <- names(params$data_files)[i]

  imports[[i]] <- import
}

# Check concenctration unit concistency over batches
if (length(unique(conc_col_name)) == 1) {
  # If consistent/unique then replace global concentration name
  assign("CONCENTRATION", unique(conc_col_name), ENV)
} else {
  # Otherwise replace concentration column names by global name without units
  for (i in seq_len(num_batches)){
    names(imports[[i]])[conc_col_name[i]] <- ENV$CONCENTRATION
  }
  # And warn user
  cat(paste0("Inconsistent concentration units between batches:\n",
                 paste0("\t", names(conc_col_name), ":\t", conc_col_name,
                       collapse = "\n"),
                 "\nUsing unit-free column name instead. But check your data!"))
}

# Merge batches
biocrates <- bind_rows(imports)

# Set "Batch" if data set contains multi batches, else NULL
if (length(unique(biocrates$Batch)) > 1) {
  BATCH <- "Batch"
  MULTIBATCH <- TRUE
} else {
  BATCH <- NULL
}

# Add batch to sample name
if (!is.null(BATCH)) {
  biocrates$Sample.Name <- paste0(biocrates$Sample.Name, "_", biocrates$Batch)
  # biocrates$Sequence.Position <- paste(sprintf("%02d", biocrates$Sequence.Position),
  #                                   biocrates$Batch)
  # biocrates$Sequence.Position <- factor(
  #   biocrates$Sequence.Position, levels = sort(unique(biocrates$Sequence.Position)))
}

# Set available data types
if (meas_type == "LC") {
  DATA_TYPES <- c(
    CONCENTRATION = ENV$CONCENTRATION,
    AREA = ENV$AREA,
    INTENSITY = ENV$INTENSITY,
    ISTD_AREA = ENV$ISTD_AREA,
    ISTD_INTENSITY = ENV$ISTD_INTENSITY
  )
} else { # "FIA"
  DATA_TYPES <- c(
    CONCENTRATION = ENV$CONCENTRATION,
    INTENSITY = ENV$INTENSITY,
    ISTD_INTENSITY = ENV$ISTD_INTENSITY
  )
}
assign("DATA_TYPES", DATA_TYPES, ENV)

# Check if a proper dataset is available
REQUIRED_METAQUAC_COLUMNS <- c(make.names(c(
  "Sample Type",
  "Sample Identification",
  "Well Position",
  "Compound",
  "Sample Name",
  "MetIDQ_Status",
  "Class",
  "Sequence Position",
  "Sample Name",
  "Batch"
)), "Well Coordinates", DATA_TYPES)

if (!all(REQUIRED_METAQUAC_COLUMNS %in% names(biocrates)) || nrow(biocrates) < 1) {
  CONTINUE <- FALSE
  message("Error: Data import failed, further processing canceled!")
  print(biocrates)
}

duplicate_samples <- get_sample_duplicates(biocrates)

message("Error: Sample duplicates with inconsistent sample information detected!")
cat("One or more samples with the same identifier (Sample Identification) but different annotations
    in other columns have been detected. This would interfere with data processing and prevent
    reliable analysis. Thus, further processing is canceled. Please use the list of duplicated
    samples provided below for guidance to correct your data.")
easy_datatable(duplicate_samples, caption = "Sample duplicates.", show_type = "statistics")
CONTINUE <- FALSE