feat(diann): Add support for DIANN 2.0

DESCRIPTION

@@ -29,7 +29,8 @@ Suggests:
     tinytest,
     covr,
     knitr,
-    rmarkdown
+    rmarkdown,
+    arrow
 Collate: 
     'clean_ProteinProspector.R'
     'clean_Metamorpheus.R'

R/clean_DIANN.R

@@ -1,62 +1,183 @@
 #' Clean raw Diann files
 #' @param msstats_object an object of class `MSstatsDIANNFiles`.
 #' @param MBR True if analysis was done with match between runs
-#' @param quantificationColumn Use 'FragmentQuantCorrected'(default) column for quantified intensities. 'FragmentQuantRaw' can be used instead.
+#' @param quantificationColumn Use 'FragmentQuantCorrected'(default) column for quantified intensities for DIANN 1.8.x.
+#' Use 'FragmentQuantRaw' for quantified intensities for DIANN 1.9.x. 
+#' Use 'auto' for quantified intensities for DIANN 2.x where each fragment intensity is a separate column, e.g. Fr0Quantity.
 #' @return data.table
 #' @importFrom stats na.omit
 #' @keywords internal
-.cleanRawDIANN = function(msstats_object, MBR = TRUE, 
-                          quantificationColumn = "FragmentQuantCorrected") {
-  dn_input = getInputFile(msstats_object, "input")
-  dn_input = data.table::as.data.table(dn_input)
-
-  if (!is.element("PrecursorMz", colnames(dn_input))) {
-    dn_input[, PrecursorMz := NA]
-  }
-  if (!is.element('FragmentInfo', colnames(dn_input))) {
-    dn_input[, FragmentInfo := NA]
-  }
-  req_cols = c('ProteinNames', 'StrippedSequence', 
-               'ModifiedSequence', 'PrecursorCharge',
-               quantificationColumn, 'QValue', 
-               'PrecursorMz', 'FragmentInfo', 'Run')
-  if (MBR) {
-    req_cols = c(req_cols, c('LibQValue', 'LibPGQValue'))
-  } else{
-    req_cols = c(req_cols, c('GlobalQValue', 'GlobalPGQValue'))
-  }
-  dn_input = dn_input[, req_cols, with = FALSE]
-  dn_input = dn_input[, lapply(.SD, function(x) unlist(tstrsplit(x, ";"))),
-                      .SDcols = c(quantificationColumn, "FragmentInfo"), 
-                      by = setdiff(colnames(dn_input), c("FragmentInfo", quantificationColumn))]
-  if (all(is.na(dn_input[["FragmentInfo"]]))) {
-    dn_input[, FragmentInfo := paste0("Frag", 1:.N),
-             by = c("ProteinNames", "ModifiedSequence", "PrecursorCharge", "Run")]
-  }
-  dn_input[, (quantificationColumn) := lapply(.SD, as.numeric), .SDcols = quantificationColumn]
-  dn_input[, FragmentIon := sub('\\^\\.\\*', '', FragmentInfo)]
-  if (any(grepl("/", dn_input$FragmentInfo))) {
-    dn_input[, ProductCharge := unlist(strsplit(FragmentInfo, split = "/"))[[1]], by = FragmentInfo]
-    dn_input[, ProductCharge := strtoi(sub("\\.\\*\\^", "", ProductCharge))]
-  } else {
-    dn_input[, ProductCharge := 1]
-  }
-  dn_input = dn_input[!grepl("NH3", FragmentIon), ]
-  dn_input = dn_input[!grepl("H2O", FragmentIon), ]
-  dn_input = na.omit(dn_input, cols = quantificationColumn)
-  data.table::setnames(dn_input, old = c('ProteinNames', 'StrippedSequence', 
-                                         'ModifiedSequence','PrecursorCharge',
-                                         quantificationColumn, 'QValue', 
-                                         'PrecursorMz', 'FragmentIon','Run', 
-                                         'ProductCharge'),
-                       new = c('ProteinName', 'PeptideSequence', 
-                               'PeptideModifiedSequence','PrecursorCharge',
-                               'Intensity', 'DetectionQValue', 
-                               'PrecursorMz', 'FragmentIon','Run',
-                               'ProductCharge'),
-                       skip_absent = TRUE)
-  dn_input[, PeptideSequence := NULL]
-  setnames(dn_input, "PeptideModifiedSequence", "PeptideSequence")
-  .logSuccess("DIANN", "clean")
-  dn_input
+.cleanRawDIANN <- function(msstats_object, MBR = TRUE, 
+                           quantificationColumn = "FragmentQuantCorrected") {
+    dn_input <- getInputFile(msstats_object, "input")
+    dn_input <- data.table::as.data.table(dn_input)
+
+    # Process quantification columns
+    quantificationColumn <- .cleanDIANNProcessQuantificationColumns(dn_input, quantificationColumn)
+
+    # Add missing columns
+    dn_input <- .cleanDIANNAddMissingColumns(dn_input)
+
+    # Select required columns
+    dn_input <- .cleanDIANNSelectRequiredColumns(dn_input, quantificationColumn, MBR)
+
+    # Split concatenated values
+    dn_input <- .cleanDIANNSplitConcatenatedValues(dn_input, quantificationColumn)
+
+    # Process fragment information
+    dn_input <- .cleanDIANNProcessFragmentInfo(dn_input, quantificationColumn)
+
+    # Clean and filter data
+    dn_input <- .cleanDIANNCleanAndFilterData(dn_input, quantificationColumn)
+
+    # Rename columns
+    dn_input <- .cleanDIANNRenameColumns(dn_input, quantificationColumn)
+
+    .logSuccess("DIANN", "clean")
+    dn_input
 }
+
+#' Process quantification columns for DIANN 2.0 format
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @return updated quantification column name
+#' @noRd
+.cleanDIANNProcessQuantificationColumns <- function(dn_input, quantificationColumn) {
+    if (quantificationColumn == "auto") {
+        fragment_columns <- grep("^Fr[0-9]+Quantity$", names(dn_input), value = TRUE)
+        if (length(fragment_columns) == 0) {
+            stop("No fragment quantification columns found. Please check your input.")
+        }
+        dn_input[, FragmentQuantCorrected := do.call(paste, c(.SD, sep = ";")),
+                 .SDcols = fragment_columns]
+        quantificationColumn <- "FragmentQuantCorrected"
+    }
+    return(quantificationColumn)
+}
+
+#' Add missing required columns
+#' @param dn_input data.table input
+#' @return data.table with missing columns added
+#' @noRd
+.cleanDIANNAddMissingColumns <- function(dn_input) {
+    if (!is.element("PrecursorMz", colnames(dn_input))) {
+        dn_input[, PrecursorMz := NA]
+    }
+    if (!is.element('FragmentInfo', colnames(dn_input))) {
+        dn_input[, FragmentInfo := NA]
+    }
+    return(dn_input)
+}
+
+#' Select required columns based on MBR setting
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @param MBR logical indicating if match between runs was used
+#' @return data.table with selected columns
+#' @noRd
+.cleanDIANNSelectRequiredColumns <- function(dn_input, quantificationColumn, MBR) {
+    base_cols <- c('ProteinNames', 'StrippedSequence', 'ModifiedSequence', 
+                   'PrecursorCharge', quantificationColumn, 'QValue', 
+                   'PrecursorMz', 'FragmentInfo', 'Run')
+
+    mbr_cols <- if (MBR) {
+        c('LibQValue', 'LibPGQValue')
+    } else {
+        c('GlobalQValue', 'GlobalPGQValue')
+    }
+
+    req_cols <- c(base_cols, mbr_cols)
+    return(dn_input[, req_cols, with = FALSE])
+}
+
+#' Split concatenated values in quantification and fragment info columns
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @return data.table with split values
+#' @noRd
+.cleanDIANNSplitConcatenatedValues <- function(dn_input, quantificationColumn) {
+    split_cols <- c(quantificationColumn, "FragmentInfo")
+    by_cols <- setdiff(colnames(dn_input), split_cols)
+
+    dn_input <- dn_input[, lapply(.SD, function(x) unlist(tstrsplit(x, ";"))),
+                         .SDcols = split_cols, 
+                         by = by_cols]
+    return(dn_input)
+}
+
+#' Process fragment information and add derived columns
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @return data.table with processed fragment info
+#' @noRd
+.cleanDIANNProcessFragmentInfo <- function(dn_input, quantificationColumn) {
+    # Generate fragment info if missing
+    if (all(is.na(dn_input[["FragmentInfo"]]))) {
+        dn_input[, FragmentInfo := paste0("Frag", 1:.N),
+                 by = c("ProteinNames", "ModifiedSequence", "PrecursorCharge", "Run")]
+    }
+
+    # Convert quantification column to numeric
+    dn_input[, (quantificationColumn) := lapply(.SD, as.numeric), 
+             .SDcols = quantificationColumn]
+
+    # Process fragment ion information
+    dn_input[, FragmentIon := sub('\\^\\.\\*', '', FragmentInfo)]
+
+    # Extract product charge
+    if (any(grepl("/", dn_input$FragmentInfo))) {
+        dn_input[, ProductCharge := .cleanDIANNExtractProductCharge(FragmentInfo), by = FragmentInfo]
+    } else {
+        dn_input[, ProductCharge := 1]
+    }
+
+    return(dn_input)
+}
+
+#' Extract product charge from fragment info
+#' @param fragment_info fragment information string
+#' @return numeric product charge
+#' @noRd
+.cleanDIANNExtractProductCharge <- function(fragment_info) {
+    charge_part <- unlist(strsplit(fragment_info, split = "/"))[[1]]
+    return(strtoi(sub("\\.\\*\\^", "", charge_part)))
+}
+
+#' Clean and filter data by removing unwanted fragments and NA values
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @return cleaned data.table
+#' @noRd
+.cleanDIANNCleanAndFilterData <- function(dn_input, quantificationColumn) {
+    # Remove NH3 and H2O loss fragments
+    dn_input <- dn_input[!grepl("NH3", FragmentIon)]
+    dn_input <- dn_input[!grepl("H2O", FragmentIon)]
+
+    # Remove rows with NA in quantification column
+    dn_input <- na.omit(dn_input, cols = quantificationColumn)
+
+    return(dn_input)
+}
+
+#' Rename columns to standardized names
+#' @param dn_input data.table input
+#' @param quantificationColumn quantification column name
+#' @return data.table with renamed columns
+#' @noRd
+.cleanDIANNRenameColumns <- function(dn_input, quantificationColumn) {
+    old_names <- c('ProteinNames', 'StrippedSequence', 'ModifiedSequence',
+                   'PrecursorCharge', quantificationColumn, 'QValue', 
+                   'PrecursorMz', 'FragmentIon', 'Run', 'ProductCharge')
+
+    new_names <- c('ProteinName', 'PeptideSequence', 'PeptideModifiedSequence',
+                   'PrecursorCharge', 'Intensity', 'DetectionQValue', 
+                   'PrecursorMz', 'FragmentIon', 'Run', 'ProductCharge')
+
+    data.table::setnames(dn_input, old = old_names, new = new_names, skip_absent = TRUE)
+
+    # Clean up peptide sequence columns
+    dn_input[, PeptideSequence := NULL]
+    setnames(dn_input, "PeptideModifiedSequence", "PeptideSequence")
+
+    return(dn_input)
+}

R/converters_DIANNtoMSstatsFormat.R

@@ -20,7 +20,9 @@
 #' @param removeFewMeasurements should proteins with few measurements be removed
 #' @param removeOxidationMpeptides should peptides with oxidation be removed
 #' @param removeProtein_with1Feature should proteins with a single feature be removed
-#' @param quantificationColumn Use 'FragmentQuantCorrected'(default) column for quantified intensities. 'FragmentQuantRaw' can be used instead.
+#' @param quantificationColumn Use 'FragmentQuantCorrected'(default) column for quantified intensities for DIANN 1.8.x.
+#' Use 'FragmentQuantRaw' for quantified intensities for DIANN 1.9.x. 
+#' Use 'auto' for quantified intensities for DIANN 2.x where each fragment intensity is a separate column, e.g. Fr0Quantity.
 #' @param ... additional parameters to `data.table::fread`.
 #'  
 #' @return data.frame in the MSstats required format.
@@ -30,7 +32,6 @@
 #' @export
 #' 
 #' @examples 
-#' # See https://github.com/vdemichev/DiaNN/discussions/1525 for workaround for DIANN 2.0
 #' input_file_path = system.file("tinytest/raw_data/DIANN/diann_input.tsv", 
 #'                                 package="MSstatsConvert")
 #' annotation_file_path = system.file("tinytest/raw_data/DIANN/annotation.csv", 
@@ -40,6 +41,17 @@
 #' output = DIANNtoMSstatsFormat(input, annotation = annot, MBR = FALSE, 
 #'                                 use_log_file = FALSE)
 #' head(output)
+#' 
+#' # For DIANN 2.0, set quantificationColumn = 'auto'
+#' input_file_path_2_0 = system.file("tinytest/raw_data/DIANN/diann_2.0.parquet", 
+#'                                 package="MSstatsConvert")
+#' annotation_file_path_2_0 = system.file("tinytest/raw_data/DIANN/annotation_diann_2.0.csv", 
+#'                                 package = "MSstatsConvert")
+#' input_2_0 = arrow::read_parquet(input_file_path_2_0)
+#' annot_2_0 = data.table::fread(annotation_file_path_2_0)
+#' output_2_0 = DIANNtoMSstatsFormat(input_2_0, annotation = annot_2_0, MBR = FALSE, 
+#'                                 use_log_file = FALSE, quantificationColumn = 'auto')
+#' head(output_2_0)
 DIANNtoMSstatsFormat = function(input, annotation = NULL,
                                 global_qvalue_cutoff = 0.01,
                                 qvalue_cutoff = 0.01, 

inst/tinytest/raw_data/DIANN/annotation_diann_2.0.csv

@@ -0,0 +1,9 @@
+Run,BioReplicate,Condition
+Run1,1,Control
+Run2,2,Control
+Run3,3,Control
+Run4,4,Control
+Run5,5,Treatment
+Run6,6,Treatment
+Run7,7,Treatment
+Run8,8,Treatment

inst/tinytest/test_converters_DIANNtoMSstatsFormat.R

@@ -16,4 +16,24 @@ expect_true("ProductCharge" %in% colnames(output))
 expect_true("IsotopeLabelType" %in% colnames(output))
 expect_true("Condition" %in% colnames(output))
 expect_true("BioReplicate" %in% colnames(output))
+expect_true("Fraction" %in% colnames(output))
+
+# Test DIANNtoMSstatsFormat DIANN 2.0 ------------------------
+input_file_path = system.file("tinytest/raw_data/DIANN/diann_2.0.parquet", package="MSstatsConvert")
+annotation_file_path = system.file("tinytest/raw_data/DIANN/annotation_diann_2.0.csv", package = "MSstatsConvert")
+input = arrow::read_parquet(input_file_path)
+annot = data.table::fread(annotation_file_path)
+output = DIANNtoMSstatsFormat(input, annotation = annot, MBR = FALSE, use_log_file = FALSE, quantificationColumn = 'auto')
+expect_equal(ncol(output), 11)
+expect_equal(nrow(output), 180)
+expect_true("Run" %in% colnames(output))
+expect_true("ProteinName" %in% colnames(output))
+expect_true("PeptideSequence" %in% colnames(output))
+expect_true("PrecursorCharge" %in% colnames(output))
+expect_true("Intensity" %in% colnames(output))
+expect_true("FragmentIon" %in% colnames(output))
+expect_true("ProductCharge" %in% colnames(output))
+expect_true("IsotopeLabelType" %in% colnames(output))
+expect_true("Condition" %in% colnames(output))
+expect_true("BioReplicate" %in% colnames(output))
 expect_true("Fraction" %in% colnames(output))