comparison preprocessing.xml @ 2:1b875f0b8024 draft

planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/cardinal commit f127be2141cf22e269c85282d226eb16fe14a9c1

1:1b22c1e7bfe7

<tool id="cardinal_preprocessing" name="MSI preprocessing" version="@VERSION@.1">
    <description>
        mass spectrometry imaging preprocessing
    </description>
    <macros>
        <import>macros.xml</import>

        @INPUT_LINKING@
        cat '${cardinal_preprocessing}' &&
        Rscript '${cardinal_preprocessing}' &&

        #if $imzml_output:
        mkdir $outfile_imzml.files_path &&
        ls -l &&
            mv ./out.imzML "${os.path.join($outfile_imzml.files_path, 'imzml')}" | true &&
            mv ./out.ibd "${os.path.join($outfile_imzml.files_path, 'ibd')}" | true &&
        #end if
            echo "imzML file:" > $outfile_imzml &&
            ls -l "$outfile_imzml.files_path" >> $outfile_imzml


    ]]>
    </command>
    <configfiles>
        <configfile name="cardinal_preprocessing"><![CDATA[

library(lattice)
library(ggplot2)

@READING_MSIDATA@


## remove duplicated coordinates, otherwise peak picking and log2 transformation will fail
print(paste0(sum(duplicated(coord(msidata))), " duplicated coordinates were removed"))
msidata <- msidata[,!duplicated(coord(msidata))]

print(paste0("Number of NA in input file: ",sum(is.na(spectra(msidata)[]))))


if (sum(spectra(msidata)[]>0, na.rm=TRUE)> 0){
    ######################### preparations for QC report #################

        maxfeatures = length(features(msidata))
        medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
        medint = round(median(spectra(msidata)[],na.rm=TRUE), digits=2)
        minmz = round(min(mz(msidata)), digits=2)
        maxmz = round(max(mz(msidata)), digits=2)
        QC_numbers= data.frame(inputdata = c(minmz, maxmz,maxfeatures, medianpeaks, medint))
        vectorofactions = "inputdata"

    ############################### Preprocessing steps ###########################
    ###############################################################################

    #for $method in $methods:

            msidata = normalize(msidata, method="tic")

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE),)
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                normalized = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, normalized)
                vectorofactions = append(vectorofactions, "normalized")

    ############################### Baseline reduction ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Baseline_reduction':
            print('Baseline_reduction')

            msidata = reduceBaseline(msidata, method="median", blocks=$method.methods_conditional.blocks_baseline, spar=$method.methods_conditional.spar_baseline)

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                baseline = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, baseline)
                vectorofactions = append(vectorofactions, "baseline red.")

    ############################### Smoothing ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Smoothing':
            print('Smoothing')

            #end if

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                smoothed = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, smoothed)
                vectorofactions = append(vectorofactions, "smoothed")

    ############################### Peak picking ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Peak_picking':
            print('Peak_picking')

            #end if

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                picked = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, picked)
                vectorofactions = append(vectorofactions, "picked")

    ############################### Peak alignment ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Peak_alignment':
            print('Peak_alignment')

           #end if

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                aligned = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, aligned)
                vectorofactions = append(vectorofactions, "aligned")

    ############################### Peak filtering ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Peak_filtering':
            print('Peak_filtering')

            msidata = peakFilter(msidata, method='freq', freq.min = $method.methods_conditional.frequ_filtering)

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                filtered = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, filtered)
                vectorofactions = append(vectorofactions, "filtered")

    ############################### Data reduction ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Data_reduction':
            print('Data_reduction')

                msidata = reduceDimension(msidata, method="bin", width=$method.methods_conditional.methods_for_reduction.bin_width, units="$method.methods_conditional.methods_for_reduction.bin_units", fun=$method.methods_conditional.methods_for_reduction.bin_fun)

                ## optional: replace NA with 0
                #if $method.methods_conditional.methods_for_reduction.replace_NA_bin:
                    print(paste0("Number of NA that were set to zero after binning:",sum(is.na(spectra(msidata)[]))))
                    spectra(msidata)[][is.na(spectra(msidata)[])] = 0
                #end if

            #elif str( $method.methods_conditional.methods_for_reduction.reduction_method) == 'resample':
                print('resample reduction')

                msidata = reduceDimension(msidata, method="peaks", ref=peak_reference, type="$method.methods_conditional.methods_for_reduction.peaks_type")
            #end if
            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                reduced = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, reduced)
                vectorofactions = append(vectorofactions, "reduced")

        ############################### Transformation ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Transformation':
            print('Transformation')

            #if str( $method.methods_conditional.transf_conditional.trans_type) == 'log2':
                print('log2 transformation')

                ## replace 0 with NA
                spectra_df = spectra(msidata)[]
                spectra_df[spectra_df ==0] = NA
                print(paste0("Number of 0 which were converted into NA:",sum(is.na(spectra_df))))
                spectra(msidata) = spectra_df
                ## log transformation

           #end if

            ############################### QC ###########################

                maxfeatures = length(features(msidata))
                medianpeaks = median(colSums(spectra(msidata)[]>0, na.rm=TRUE))
                medint = round(median(spectra(msidata)[], na.rm=TRUE), digits=2)
                minmz = round(min(mz(msidata)), digits=2)
                maxmz = round(max(mz(msidata)), digits=2)
                transformed = c(minmz, maxmz,maxfeatures, medianpeaks, medint)
                QC_numbers= cbind(QC_numbers, transformed)
                vectorofactions = append(vectorofactions, "transformed")

            #end if
    #end for

    ############# Outputs: RData, imzml and QC report #############
    ################################################################################

    print(paste0("Number of NA in output file: ",sum(is.na(spectra(msidata)[]))))

    ## save as (.RData)
    save(msidata, file="$msidata_preprocessed")

    ## save msidata as imzML file, will only work if there is at least 1 m/z left
    #if $imzml_output:
        if (nrow(msidata) > 0){
print("write outputfile")
            writeImzML(msidata, "out")}
    #end if

    ## save QC report

    pdf("Preprocessing.pdf", fonts = "Times", pointsize = 12)
    plot(0,type='n',axes=FALSE,ann=FALSE)
    title(main=paste("Quality control during preprocessing \n", "Filename:", "$infile.display_name"))
    rownames(QC_numbers) = c("min m/z", "max mz", "# features", "median\n# peaks", "median\nintensity")
    grid.table(t(QC_numbers))

    dev.off()

}else{

    <inputs>
        <expand macro="reading_msidata"/>
        <repeat name="methods" title="Preprocessing" min="1" max="50">
            <conditional name="methods_conditional">
                <param name="preprocessing_method" type="select" label="Preprocessing methods">
                    <option value="Normalization" selected="True">Normalization</option>
                    <option value="Baseline_reduction">Baseline Reduction</option>
                    <option value="Smoothing">Peak smoothing</option>
                    <option value="Peak_picking">Peak picking</option>
                    <option value="Peak_alignment">Peak alignment</option>
                    <option value="Peak_filtering">Peak filtering</option>

                            <when value="sqrt"/>
                    </conditional>
                </when>
            </conditional>
        </repeat>
        <param name="imzml_output" type="boolean" label="Output of imzML file" truevalue="TRUE" falsevalue="FALSE"/>
    </inputs>
    <outputs>
        <data format="rdata" name="msidata_preprocessed" label="${tool.name} on ${on_string}"/>
        <data format="pdf" name="QC_overview" from_work_dir="Preprocessing.pdf" label = "${tool.name} on ${on_string}: QC"/>
        <data format="imzml" name="outfile_imzml" label="${tool.name} on ${on_string}: imzML">
            <filter>imzml_output</filter>
       </data>
    </outputs>
    <tests>
        <test>
            <expand macro="infile_imzml"/>
            <repeat name="methods">

                        <conditional name="transf_conditional">
                            <param name="trans_type" value="sqrt"/>
                        </conditional>
                </conditional>
            </repeat>
            <output name="msidata_preprocessed" file="preprocessing_results1.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results1.pdf" compare="sim_size"/>
        </test>
        <test>
            <param name="infile" value="3_files_combined.RData" ftype="rdata"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                    <conditional name="methods_for_alignment">
                        <param name="alignment_method" value="DP"/>
                    </conditional>
                </conditional>
            </repeat>
            <output name="msidata_preprocessed" file="preprocessing_results2.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results2.pdf" compare="sim_size"/>
        </test>
        <test>
            <expand macro="infile_analyze75"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                    <conditional name="methods_for_alignment">
                        <param name="alignment_method" value="diff"/>
                    </conditional>
                </conditional>
            </repeat>
            <output name="msidata_preprocessed" file="preprocessing_results3.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results3.pdf" compare="sim_size"/>
        </test>
        <test>
            <expand macro="infile_analyze75"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                <conditional name="methods_conditional">
                    <param name="preprocessing_method" value="Data_reduction"/>
                    <param name="bin_width" value="0.1"/>
                </conditional>
            </repeat>
            <output name="msidata_preprocessed" file="preprocessing_results4.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results4.pdf" compare="sim_size"/>
        </test>
        <test>
            <expand macro="infile_imzml"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                            <param name="reduction_method" value="resample"/>
                            <param name="step_width" value="0.1"/>
                        </conditional>
                </conditional>
            </repeat>
             <output name="msidata_preprocessed" file="preprocessing_results5.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results5.pdf" compare="sim_size"/>
        </test>
    </tests>
    <help>
        <![CDATA[

- Peak alignment works only after peak picking
- Peak filtering works only on centroided data (peak picking and alignment or Data reduction peaks

**Output**

- MSI data as .RData output (can be read with the Cardinal package in R)
- optional: MSI data as imzML file
- pdf with key values after each processing step

        ]]>
    </help>
    <expand macro="citations"/>
</tool>

2:1b875f0b8024

<tool id="cardinal_preprocessing" name="MSI preprocessing" version="@VERSION@.2">
    <description>
        mass spectrometry imaging preprocessing
    </description>
    <macros>
        <import>macros.xml</import>

        @INPUT_LINKING@
        cat '${cardinal_preprocessing}' &&
        Rscript '${cardinal_preprocessing}' &&

        #if str($imzml_output) == "imzml_format":
        mkdir $outfile_imzml.files_path &&
            mv ./out.imzML "${os.path.join($outfile_imzml.files_path, 'imzml')}" | true &&
            mv ./out.ibd "${os.path.join($outfile_imzml.files_path, 'ibd')}" | true &&
        #end if
        echo "imzML file:" > $outfile_imzml &&
        ls -l "$outfile_imzml.files_path" >> $outfile_imzml

    ]]>
    </command>
    <configfiles>
        <configfile name="cardinal_preprocessing"><![CDATA[

library(lattice)
library(ggplot2)

@READING_MSIDATA@

## remove duplicated coordinates, otherwise peak picking and log2 transformation will fail
print(paste0(sum(duplicated(coord(msidata)[,1:2])), " duplicated coordinates were removed"))
msidata <- msidata[,!duplicated(coord(msidata)[,1:2])]


if (ncol(msidata)>0 & nrow(msidata) >0){                                         

    ## start QC report

    pdf("Preprocessing.pdf", fonts = "Times", pointsize = 12)
    plot(0,type='n',axes=FALSE,ann=FALSE)
    title(main=paste("Quality control during preprocessing \n", "Filename:", "$infile.display_name"))

    ######################### preparations for QC report #################

    maxfeatures =nrow(msidata)
    pixelcount = ncol(msidata)
    minmz = round(min(mz(msidata)), digits=2)
    maxmz = round(max(mz(msidata)), digits=2)
    QC_numbers= data.frame(inputdata = c(minmz, maxmz,maxfeatures, pixelcount))
    vectorofactions = "inputdata"
    plot(msidata, pixel = 1:pixelcount, main="Average spectrum of input file")

    ############################### Preprocessing steps ###########################
    ###############################################################################

    #for $method in $methods:

            msidata = normalize(msidata, method="tic")

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            normalized = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, normalized)
            vectorofactions = append(vectorofactions, "normalized")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after normalization")

    ############################### Baseline reduction ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Baseline_reduction':
            print('Baseline_reduction')

            msidata = reduceBaseline(msidata, method="median", blocks=$method.methods_conditional.blocks_baseline, spar=$method.methods_conditional.spar_baseline)

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            baseline = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, baseline)
            vectorofactions = append(vectorofactions, "baseline red.")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after baseline reduction")

    ############################### Smoothing ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Smoothing':
            print('Smoothing')

            #end if

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            smoothed = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, smoothed)
            vectorofactions = append(vectorofactions, "smoothed")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after smoothing")

    ############################### Peak picking ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Peak_picking':
            print('Peak_picking')

            #end if

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            picked = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, picked)
            vectorofactions = append(vectorofactions, "picked")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after peak picking")

    ############################### Peak alignment ###########################

        #elif str( $method.methods_conditional.preprocessing_method ) == 'Peak_alignment':
            print('Peak_alignment')

           #end if

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            aligned = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, aligned)
            vectorofactions = append(vectorofactions, "aligned")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after alignment")

    ############################### Peak filtering ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Peak_filtering':
            print('Peak_filtering')

            msidata = peakFilter(msidata, method='freq', freq.min = $method.methods_conditional.frequ_filtering)

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            filtered = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, filtered)
            vectorofactions = append(vectorofactions, "filtered")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after filtering")

    ############################### Data reduction ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Data_reduction':
            print('Data_reduction')

                msidata = reduceDimension(msidata, method="bin", width=$method.methods_conditional.methods_for_reduction.bin_width, units="$method.methods_conditional.methods_for_reduction.bin_units", fun=$method.methods_conditional.methods_for_reduction.bin_fun)

                ## optional: replace NA with 0
                #if $method.methods_conditional.methods_for_reduction.replace_NA_bin:
                    ## binning seems to create normal R matrix but to be sure:
                    iData(msidata) <- iData(msidata)[]
                    ## count and replace NAs
                    print(paste0("Number of NA that were set to zero after binning:",sum(is.na(spectra(msidata)[]))))
                    spectra(msidata)[is.na(spectra(msidata))] = 0 
                #end if

            #elif str( $method.methods_conditional.methods_for_reduction.reduction_method) == 'resample':
                print('resample reduction')

                msidata = reduceDimension(msidata, method="peaks", ref=peak_reference, type="$method.methods_conditional.methods_for_reduction.peaks_type")
            #end if
            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            reduced = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, reduced)
            vectorofactions = append(vectorofactions, "reduced")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after data reduction")

        ############################### Transformation ###########################

        #elif str( $method.methods_conditional.preprocessing_method) == 'Transformation':
            print('Transformation')

            ## convert data into R matrix what brings it automatically into memory and can take some take but next steps need R matrix
            iData(msidata) <- iData(msidata)[]

            #if str( $method.methods_conditional.transf_conditional.trans_type) == 'log2':
                print('log2 transformation')

                ## replace 0 with NA to prevent Inf
                spectra_df = spectra(msidata)[]
                spectra_df[spectra_df ==0] = NA
                print(paste0("Number of 0 which were converted into NA:",sum(is.na(spectra_df))))
                spectra(msidata) = spectra_df
                ## log transformation

           #end if

            ############################### QC ###########################

            maxfeatures =nrow(msidata)
            pixelcount = ncol(msidata)
            minmz = round(min(mz(msidata)), digits=2)
            maxmz = round(max(mz(msidata)), digits=2)
            transformed = c(minmz, maxmz,maxfeatures, pixelcount)
            QC_numbers= cbind(QC_numbers, transformed)
            vectorofactions = append(vectorofactions, "transformed")
            plot(msidata, pixel = 1:pixelcount, main="Average spectrum after transformation")

            #end if
    #end for

    ############# Outputs: RData, imzml and QC report #############
    ################################################################################

    ## save msidata as imzML file, will only work if there is at least 1 m/z left

    #if str($imzml_output) == "imzml_format":
        if (nrow(msidata) > 0){
            writeImzML(msidata, "out")}
    #elif str($imzml_output) == "rdata_format":
        ## save as (.RData)
        iData(msidata) = iData(msidata)[]
        save(msidata, file="$outfile_rdata")
    #end if

    plot(0,type='n',axes=FALSE,ann=FALSE)
    rownames(QC_numbers) = c("min m/z", "max mz", "# features", "# spectra")
    grid.table(t(QC_numbers))

    dev.off()

}else{

    <inputs>
        <expand macro="reading_msidata"/>
        <repeat name="methods" title="Preprocessing" min="1" max="50">
            <conditional name="methods_conditional">
                <param name="preprocessing_method" type="select" label="Preprocessing methods">
                    <option value="Normalization" selected="True">Intensity Normalization (TIC)</option>
                    <option value="Baseline_reduction">Baseline Reduction</option>
                    <option value="Smoothing">Peak smoothing</option>
                    <option value="Peak_picking">Peak picking</option>
                    <option value="Peak_alignment">Peak alignment</option>
                    <option value="Peak_filtering">Peak filtering</option>

                            <when value="sqrt"/>
                    </conditional>
                </when>
            </conditional>
        </repeat>
        <param name="imzml_output" type="select" display = "radio" optional = "False"
               label="Output format" help= "Choose the output format">
                <option value="imzml_format" selected="True">imzML</option>
                <option value="rdata_format">RData</option>
        </param>
    </inputs>
    <outputs>
        <data format="imzml" name="outfile_imzml" label="${tool.name} on ${on_string}: imzML">
            <filter>imzml_output=='imzml_format'</filter>
        </data>
        <data format="rdata" name="outfile_rdata" label="${tool.name} on ${on_string}: RData">
            <filter>imzml_output == 'rdata_format'</filter>
        </data>
        <data format="pdf" name="QC_overview" from_work_dir="Preprocessing.pdf" label = "${tool.name} on ${on_string}: QC"/>
    </outputs>
    <tests>
        <test>
            <expand macro="infile_imzml"/>
            <repeat name="methods">

                        <conditional name="transf_conditional">
                            <param name="trans_type" value="sqrt"/>
                        </conditional>
                </conditional>
            </repeat>
            <param name="imzml_output" value="imzml_format"/>
            <output name="QC_overview" file="preprocessing_results1.pdf" compare="sim_size"/>
            <output name="outfile_imzml" ftype="imzml" file="preprocessing_results1.imzml.txt" compare="sim_size">
                <extra_files type="file" file="preprocessing_results1.imzml" name="imzml" lines_diff="4"/>
                <extra_files type="file" file="preprocessing_results1.ibd" name="ibd" compare="sim_size"/>
            </output>
        </test>
        <test>
            <param name="infile" value="3_files_combined.RData" ftype="rdata"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                    <conditional name="methods_for_alignment">
                        <param name="alignment_method" value="DP"/>
                    </conditional>
                </conditional>
            </repeat>
            <param name="imzml_output" value="imzml_format"/>
            <output name="QC_overview" file="preprocessing_results2.pdf" compare="sim_size"/>
            <output name="outfile_imzml" ftype="imzml" file="preprocessing_results2.imzml.txt" compare="sim_size">
                <extra_files type="file" file="preprocessing_results2.imzml" name="imzml" lines_diff="4"/>
                <extra_files type="file" file="preprocessing_results2.ibd" name="ibd" compare="sim_size"/>
            </output>
        </test>
        <test>
            <expand macro="infile_analyze75"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                    <conditional name="methods_for_alignment">
                        <param name="alignment_method" value="diff"/>
                    </conditional>
                </conditional>
            </repeat>
            <param name="imzml_output" value="imzml_format"/>
            <output name="QC_overview" file="preprocessing_results3.pdf" compare="sim_size"/>
            <output name="outfile_imzml" ftype="imzml" file="preprocessing_results3.imzml.txt" compare="sim_size">
                <extra_files type="file" file="preprocessing_results3.imzml" name="imzml" lines_diff="4"/>
                <extra_files type="file" file="preprocessing_results3.ibd" name="ibd" compare="sim_size"/>
            </output>
        </test>
        <test>
            <expand macro="infile_analyze75"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                <conditional name="methods_conditional">
                    <param name="preprocessing_method" value="Data_reduction"/>
                    <param name="bin_width" value="0.1"/>
                </conditional>
            </repeat>
            <param name="imzml_output" value="imzml_format"/>
            <output name="QC_overview" file="preprocessing_results4.pdf" compare="sim_size"/>
            <output name="outfile_imzml" ftype="imzml" file="preprocessing_results4.imzml.txt" compare="sim_size">
                <extra_files type="file" file="preprocessing_results4.imzml" name="imzml" lines_diff="4"/>
                <extra_files type="file" file="preprocessing_results4.ibd" name="ibd" compare="sim_size"/>
            </output>
        </test>
        <test>
            <expand macro="infile_imzml"/>
            <repeat name="methods">
                <conditional name="methods_conditional">

                            <param name="reduction_method" value="resample"/>
                            <param name="step_width" value="0.1"/>
                        </conditional>
                </conditional>
            </repeat>
            <param name="imzml_output" value="rdata_format"/>
            <output name="outfile_rdata" file="preprocessing_results5.RData" compare="sim_size"/>
            <output name="QC_overview" file="preprocessing_results5.pdf" compare="sim_size"/>
        </test>
    </tests>
    <help>
        <![CDATA[

- Peak alignment works only after peak picking
- Peak filtering works only on centroided data (peak picking and alignment or Data reduction peaks

**Output**

- MSI data as imzML file or .RData (can be read with the Cardinal package in R)
- pdf with key values and average mass spectra after each processing step

        ]]>
    </help>
    <expand macro="citations"/>
</tool>