diff data_exporter.xml @ 7:350a84ea795c draft

"planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/cardinal commit f986c51abe33c7f622d429a3c4a79ee24b33c1f3"

--- a/data_exporter.xml	Wed Mar 25 08:07:01 2020 -0400
+++ b/data_exporter.xml	Thu Apr 23 08:06:24 2020 -0400
@@ -1,4 +1,4 @@
-<tool id="cardinal_data_exporter" name="MSI data exporter" version="@VERSION@.2">
+<tool id="cardinal_data_exporter" name="MSI data exporter" version="@VERSION@.0">
     <description>
         exports imzML and Analyze7.5 to tabular files
     </description>
@@ -22,22 +22,27 @@
 
 library(Cardinal)
 
-@READING_MSIDATA_INRAM@
+@READING_MSIDATA@
 
+## in case RData input is MSImageSet:
+if (class(msidata) == "MSImageSet"){
+    msidata = as(msidata, "MSImagingExperiment")
+    run(msidata) = "infile"
+    }
 
+    ## extract spectra matrix once: 
+    msidata_matrix = as.matrix(iData(msidata))
     ###################### Intensity matrix output ################################
 
     #if "int_matrix" in str($output_options).split(","):
         print("intensity matrix output")
 
-        mz_names = gsub(" = ", "_", names(features(msidata)))
-        mz_names = gsub("/", "", mz_names)
-        pixel_names = gsub(", y = ", "_", names(pixels(msidata)))
-        pixel_names = gsub(" = ", "y_", pixel_names)
+        mz_names = paste0("mz_", mz(msidata))
+        pixel_names = paste0("xy_", coord(msidata)\$x, "_", coord(msidata)\$y)
 
-        ##spectramatrix = cbind(mz_names,spectra(msidata))
-        newmatrix = rbind(c("mz_name", pixel_names), cbind(mz_names,spectra(msidata)))
-        write.table(newmatrix, file="$intensity_matrix", quote = FALSE, row.names = FALSE, col.names=FALSE, sep = "\t")
+        write.table(
+                    rbind(c("mz_name", pixel_names), cbind(mz_names,msidata_matrix)), ##create matrix
+                    file="$intensity_matrix", quote = FALSE, row.names = FALSE, col.names=FALSE, sep = "\t")
         ## free up RAM space in case furhter steps will be run: 
         rm(newmatrix)
         gc()
@@ -49,21 +54,18 @@
     #if "mz_tabular" in str($output_options).split(","):
         print("mz feature output")
 
-        mz_names = gsub(" = ", "_", names(features(msidata)))
-        mz_names = gsub("/", "", mz_names)
+        full_sample_mean = apply(msidata_matrix,1,mean, na.rm=TRUE)
+        full_sample_sd = apply(msidata_matrix,1,sd, na.rm=TRUE)
 
-        ## mean, median, sd and SEM intensity per file and mz
-        full_sample_mean = rowMeans(spectra(msidata), na.rm=TRUE)
-        full_sample_median = apply(spectra(msidata),1,median, na.rm=TRUE)
-        full_sample_sd = apply(spectra(msidata),1,sd, na.rm=TRUE)
-        full_sample_sem = full_sample_sd/full_sample_mean*100
-        ## npeaks and sum of all intensities per spectrum and mz
-        npeaks= sum(spectra(msidata)>0, na.rm=TRUE)
-        mzTIC = rowSums(spectra(msidata), na.rm=TRUE) ## calculate intensity sum for each m/z
-        peakspermz = rowSums(spectra(msidata) > 0, na.rm=TRUE) ## calculate number of intensities > 0 for each m/z (max = number of spectra)
+        mz_df = data.frame(paste0("mz_", mz(msidata)),                 ##mz names
+                           mz(msidata),                                ##mz values
+                           full_sample_mean,                           ##mean
+                           apply(msidata_matrix,1,median, na.rm=TRUE), ##median
+                           full_sample_sd,                             ##sd
+                           full_sample_sd/full_sample_mean*100,        ##SEM
+                           rowSums(msidata_matrix, na.rm=TRUE),        ## calculate intensity sum for each m/z
+                           rowSums(msidata_matrix > 0, na.rm=TRUE))    ## calculate number of intensities > 0 for each m/z (max = number of spectra)
 
-        ## combine into dataframe, order is the same for all vectors
-        mz_df = data.frame(mz_names, mz(msidata), full_sample_mean, full_sample_median, full_sample_sd, full_sample_sem, mzTIC, peakspermz)
         colnames(mz_df) = c("mz_names", "mz", "sample_mean", "sample_median", "sample_sd", "sample_sem", "intensity_sum", "number_peaks")
         write.table(mz_df, file="$feature_output", quote = FALSE, row.names = FALSE, col.names=TRUE, sep = "\t")
         ## free up RAM space in case furhter steps will be run: 
@@ -82,16 +84,15 @@
         colnames(annotation_input) = c("x", "y", "annotation")
 
         ## merge with coordinate information of msidata
-        msidata_coordinates = cbind(coord(msidata)[,1:2], c(1:ncol(msidata)))
-        colnames(msidata_coordinates)[3] = "pixel_index"
+        msidata_coordinates = data.frame(coord(msidata)\$x, coord(msidata)\$y, c(1:ncol(msidata)))
+        colnames(msidata_coordinates) = c("x", "y", "pixel_index")
         merged_annotation = merge(msidata_coordinates, annotation_input, by=c("x", "y"), all.x=TRUE)
         merged_annotation[is.na(merged_annotation)] = "NA"
         merged_annotation = merged_annotation[order(merged_annotation\$pixel_index),]
         msidata\$annotation = as.factor(merged_annotation[,4])
 
         ## create m/z feature name
-        mz_names = gsub(" = ", "_", names(features(msidata)))
-        mz_names = gsub("/", "", mz_names)
+        mz_names = paste0("mz_", mz(msidata))
 
         #if "mean" in str($tabular_annotation.summary_type).split(","):
             print("summarized mean")
@@ -131,7 +132,7 @@
             count = 1
             for (subsample in levels(msidata\$annotation)){
                 subsample_pixels = msidata[,msidata\$annotation == subsample]
-                subsample_calc = apply(spectra(subsample_pixels),1,sd, na.rm=TRUE)
+                subsample_calc = apply(as.matrix(spectra(subsample_pixels)),1,sd, na.rm=TRUE)
                 sample_matrix = cbind(sample_matrix, subsample_calc)
                 count = count+1}
             sample_matrix_sd = cbind(mz_names,sample_matrix)
@@ -147,40 +148,32 @@
         print("pixel output")
 
         ## coordinates
-        xycoordinates = coord(msidata)[,1:2]
+        xycoordinates = data.frame(coord(msidata)\$x, coord(msidata)\$y)
+        colnames(xycoordinates) = c("x", "y")
 
         ## pixel name
-        pixel_names = paste0("xy_", xycoordinates\$x, "_", xycoordinates\$y)
-
-        ## pixel order 
-        pixelxyarray=1:length(pixels(msidata))
-
-        ## number of pixels per spectrum: every intensity value > 0 counts as peak
-        peaksperpixel = colSums(spectra(msidata)>0, na.rm=TRUE)
-
-        ## Total ion chromatogram per spectrum
-        TICs = round(colSums(spectra(msidata), na.rm=TRUE), digits = 2)
-
-        ## Median ion intensity per spectrum
-        med_int = round(apply(spectra(msidata), 2, median, na.rm=TRUE), digits = 2)
-
-        ## Maximum ion intensity per spectrum
-        max_int = round(apply(spectra(msidata), 2, max, na.rm=TRUE), digits = 2)
-
-        ## Highest m/z per spectrum
-        highestmz = apply(spectra(msidata),2,which.max) 
-        highestmz_data = mz(msidata)[highestmz]
+        pixel_names = paste0("xy_", coord(msidata)\$x, "_", coord(msidata)\$y)
 
         ## Combine into dataframe; order is the same for all vectors
-        spectra_df = data.frame(pixel_names, xycoordinates, pixelxyarray, peaksperpixel, med_int, TICs, max_int, highestmz_data)
+        spectra_df = data.frame(pixel_names, 
+                                xycoordinates, 
+                                1:length(pixels(msidata)),                                        ##pixel order
+                                colSums(msidata_matrix>0, na.rm=TRUE),                            ##peaks per pixel
+                                round(apply(msidata_matrix, 2, median, na.rm=TRUE), digits = 2),  ## median intensity
+                                round(colSums(msidata_matrix, na.rm=TRUE), digits = 2),           ##TICs
+                                round(apply(msidata_matrix, 2, max, na.rm=TRUE), digits = 2),     ##max intensity
+                                mz(msidata)[apply(msidata_matrix,2,which.max) ])                  ##highest mz
+
         colnames(spectra_df) = c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz")
 
         #if str($counting_calibrants.pixel_with_calibrants) == "yes_calibrants":
 
             calibrant_list = read.delim("$counting_calibrants.mz_tabular", header = $counting_calibrants.feature_header, na.strings=c("","NA"), stringsAsFactors = FALSE)
             calibrant_list = calibrant_list[,$counting_calibrants.feature_column, drop=FALSE]
+
             ### calculate how many input calibrant m/z are valid: 
             inputcalibrants = calibrant_list[calibrant_list[,1]>min(mz(msidata)) & calibrant_list[,1]<max(mz(msidata)),,drop = FALSE]
+
             inputcalibrantmasses = inputcalibrants[,1]
 
             ##QC plot number 2) Number of calibrants per spectrum
@@ -214,11 +207,11 @@
 
             }else{countvector = rep(0,ncol(msidata))}
                 countdf= cbind(coord(msidata)[,1:2], countvector) ## add pixel coordinates to counts
-                colnames(countdf) = c("x_values", "y_values", "m/z count")
+                colnames(countdf) = c("x_values", "y_values", "mz_count")
                 spectra_df = merge(spectra_df, countdf, by=c("x_values", "y_values"))
 
                 ## sort columns to have spectra_names as rowname in first column
-                spectra_df = spectra_df[c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz", "m/z count")]
+                spectra_df = spectra_df[c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz", "mz_count")]
 
         #end if
         #if str($tabular_annotation.load_annotation) == 'yes_annotation':
@@ -228,7 +221,7 @@
 
             ## sort columns to have spectra_names as rowname in first column
             #if str($counting_calibrants.pixel_with_calibrants) == "yes_calibrants":
-                spectra_df = spectra_df[c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz", "m/z count", "annotation")]
+                spectra_df = spectra_df[c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz", "mz_count", "annotation")]
             #else
                 spectra_df = spectra_df[c("spectra_names", "x_values", "y_values","pixel_order", "peaks_per_spectrum", "median_intensity", "spectrum_TIC", "maximum_intensity", "most_abundant_mz", "annotation")]
             #end if
@@ -335,6 +328,24 @@
             <output name="feature_output" file="features_out3.tabular"/>
             <output name="pixel_output" file="pixel_out3.tabular"/>
         </test>
+        <test expect_num_outputs="2">
+            <expand macro="processed_infile_imzml"/>
+            <conditional name="processed_cond">
+                <param name="processed_file" value="processed"/>
+                <param name="accuracy" value="100"/>
+                <param name="units" value="ppm"/>
+            </conditional>
+            <param name="output_options" value="pixel_tabular,mz_tabular"/>
+            <conditional name="counting_calibrants">
+                <param name="pixel_with_calibrants" value="yes_calibrants"/>
+                <param name="mz_tabular" ftype="tabular" value = "featuresofinterest5.tabular"/>
+                <param name="feature_column" value="1"/>
+                <param name="feature_header" value="False"/>
+                <param name="plusminus_ppm" value="200"/>
+            </conditional>
+            <output name="feature_output" file="features_out4.tabular"/>
+            <output name="pixel_output" file="pixel_out4.tabular"/>
+        </test>
     </tests>
     <help>
         <![CDATA[