diff --git a/tools/waveica/macros.xml b/tools/waveica/macros.xml index 551d1321..e5f9bbaf 100644 --- a/tools/waveica/macros.xml +++ b/tools/waveica/macros.xml @@ -96,10 +96,20 @@ + + + - + + + + + + + keep_two_output diff --git a/tools/waveica/test-data/test10_output1.tsv b/tools/waveica/test-data/test10_output1.tsv new file mode 100644 index 00000000..375e65d3 --- /dev/null +++ b/tools/waveica/test-data/test10_output1.tsv @@ -0,0 +1,5 @@ +id VT_160120_002 VT_160120_004 VT_160120_006 VT_160120_008 VT_160120_010 +M85T34 355200.506508035 216897.826587868 362337.195084504 143303.377379009 189065.516447239 +M86T41 75115889.9077485 75204863.1495248 76490295.1450204 83771659.9549148 84108898.7658797 +M86T518 6101488.54615418 6170882.26270475 12588041.969092 6181538.46316058 6103964.42378424 +M86T539 2007379.02604984 2069979.64992079 1818589.63912375 1975712.25920485 1935671.32085241 diff --git a/tools/waveica/test-data/test10_output2.tsv b/tools/waveica/test-data/test10_output2.tsv new file mode 100644 index 00000000..21ccd260 --- /dev/null +++ b/tools/waveica/test-data/test10_output2.tsv @@ -0,0 +1,6 @@ +sampleName class sampleType injectionOrder batch +VT_160120_002 sample sample 1 1 +VT_160120_004 sample sample 2 1 +VT_160120_006 sample sample 3 1 +VT_160120_008 sample sample 4 1 +VT_160120_010 sample sample 5 1 diff --git a/tools/waveica/test-data/test9_output1.parquet b/tools/waveica/test-data/test9_output1.parquet new file mode 100644 index 00000000..45cff33e Binary files /dev/null and b/tools/waveica/test-data/test9_output1.parquet differ diff --git a/tools/waveica/test-data/test9_output2.parquet b/tools/waveica/test-data/test9_output2.parquet new file mode 100644 index 00000000..7f1bf055 Binary files /dev/null and b/tools/waveica/test-data/test9_output2.parquet differ diff --git a/tools/waveica/waveica.xml b/tools/waveica/waveica.xml index a9f09b8f..d8756e76 100644 --- a/tools/waveica/waveica.xml +++ b/tools/waveica/waveica.xml @@ -1,4 +1,4 @@ - + removal of batch effects for untargeted metabolomics data macros.xml @@ -51,14 +51,14 @@ )' #end if - -e 'store_data(normalized_data, "$normalized_data", "$input_num.data.ext")' + -e 'store_data(normalized_data, "$normalized_data", "$metadata", "$input_num.data.ext", $keep_two_output)' ]]> - - + + @@ -89,12 +89,13 @@ + - + @@ -103,9 +104,9 @@ - + - + @@ -114,9 +115,9 @@ - + - + @@ -127,7 +128,7 @@ - + @@ -138,9 +139,9 @@ - + - + @@ -151,9 +152,9 @@ - + - + @@ -166,7 +167,7 @@ - + @@ -180,7 +181,7 @@ - + @@ -192,7 +193,7 @@ - + - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + - + diff --git a/tools/waveica/waveica_wrapper.R b/tools/waveica/waveica_wrapper.R index 1332919d..875982ac 100644 --- a/tools/waveica/waveica_wrapper.R +++ b/tools/waveica/waveica_wrapper.R @@ -3,10 +3,7 @@ read_file <- function(file, metadata, ft_ext, mt_ext, transpose) { if (transpose) { col_names <- c("sampleName", data[[1]]) - t_data <- data[-1] - t_data <- t(t_data) - data <- data.frame(rownames(t_data), t_data) - colnames(data) <- col_names + data <- tranpose_data(data, col_names) } if (!is.na(metadata)) { @@ -133,7 +130,6 @@ waveica_singlebatch <- function(file, return(data) } - sort_by_injection_order <- function(data) { if ("batch" %in% colnames(data)) { data <- data[order(data[, "batch"], data[, "injectionOrder"], decreasing = FALSE), ] @@ -143,7 +139,6 @@ sort_by_injection_order <- function(data) { return(data) } - verify_input_dataframe <- function(data, required_columns) { if (anyNA(data)) { stop("Error: dataframe cannot contain NULL values! @@ -194,7 +189,6 @@ verify_column_types <- function(data, required_columns) { return(data) } - # Match group labels with [blank/sample/qc] and enumerate them enumerate_groups <- function(group) { group[grepl("blank", tolower(group))] <- 0 @@ -204,7 +198,6 @@ enumerate_groups <- function(group) { return(group) } - # Create appropriate input for R wavelets function get_wf <- function(wavelet_filter, wavelet_length) { wf <- paste(wavelet_filter, wavelet_length, sep = "") @@ -217,7 +210,6 @@ get_wf <- function(wavelet_filter, wavelet_length) { return(wf) } - # Exclude blanks from a dataframe exclude_group <- function(data, group) { row_idx_to_exclude <- which(group %in% 0) @@ -230,14 +222,62 @@ exclude_group <- function(data, group) { } } -store_data <- function(data, output, ext) { +store_data <- function(data, feature_output, metadata_output, ext, split_output = FALSE) { if (ext == "parquet") { - arrow::write_parquet(data, output) + if (split_output == TRUE) { + split_df <- split_output(data) + arrow::write_parquet(split_df$metadata, metadata_output) + arrow::write_parquet(split_df$feature_table, feature_output) + } else { + arrow::write_parquet(data, feature_output) + } } else { - write.table(data, - file = output, sep = "\t", - row.names = FALSE, quote = FALSE - ) + if (split_output == TRUE) { + split_df <- split_output(data) + write.table(split_df$metadata, + file = metadata_output, sep = "\t", + row.names = FALSE, quote = FALSE + ) + write.table(split_df$feature_table, + file = feature_output, sep = "\t", + row.names = FALSE, quote = FALSE + ) + } else { + write.table(data, + file = feature_output, sep = "\t", + row.names = FALSE, quote = FALSE + ) + } } cat("Normalization has been completed.\n") } + +split_output <- function(df) { + required_columns_set1 <- c("sampleName", "class", "sampleType", "injectionOrder", "batch") + required_columns_set2 <- c("sampleName", "class", "sampleType", "injectionOrder") + + if (all(required_columns_set1 %in% colnames(df))) { + metadata_df <- df[, required_columns_set1, drop = FALSE] + df <- df[, -c(2:5)] + } else if (all(required_columns_set2 %in% colnames(df))) { + metadata_df <- df[, required_columns_set2, drop = FALSE] + df <- df[, -c(2:4)] + } else { + stop("Neither set of required columns is present in the dataframe.") + } + + # Transpose the feature table + col_names <- c("id", as.vector(df[[1]])) + feature_table <- tranpose_data(df, col_names) + + return(list(metadata = metadata_df, feature_table = feature_table)) +} + +tranpose_data <- function(data, column_names) { + t_data <- data[-1] + t_data <- t(t_data) + tranposed_data <- data.frame(rownames(t_data), t_data) + colnames(tranposed_data) <- column_names + + return(tranposed_data) +}