annotate bioconductor_scp.xml @ 0:cd2f3a280463 draft default tip

planemo upload for repository https://github.com/RECETOX/galaxytools/tree/master/tools/bioconductor-scp commit a0a1a3de5dd24b2aabe96ec3d6f89acdcf5e462b

<tool id="bioconductor_scp" name="bioconductor-scp" version="@TOOL_VERSION@+galaxy0" profile="23.0">
    <description>single cell proteomics data analysis workflow</description>
    <macros>
        <import>macros.xml</import>
        <import>help.xml</import>
    </macros>
    <xrefs>
        <xref type="bio.tools">scp</xref>
    </xrefs>
    <requirements>
        <requirement type="package" version="@TOOL_VERSION@">bioconductor-scp</requirement>
        <requirement type="package" version="3.54.0">bioconductor-sva</requirement>
        <requirement type="package" version="1.80.0">bioconductor-impute</requirement>
        <requirement type="package" version="1.34.0">bioconductor-scater</requirement>
        <requirement type="package" version="1.16.0">bioconductor-qfeatures</requirement>
        <requirement type="package" version="3.62.1">bioconductor-limma</requirement>
        <requirement type="package" version="3.5.1">r-ggplot2</requirement>
    </requirements>
    <required_files>
        <include path="utils.r" />
    </required_files>
    <expand macro="creator" />
    <command detect_errors="exit_code"><![CDATA[
    echo ${run_script} &&
    Rscript  -e 'source("${__tool_directory__}/utils.r")' -e 'source("${run_script}")'
    #if $data_export.export_R_script
    && cat ${run_script} >> $script
    #end if
    ]]></command>
    <configfiles>
        <configfile name="run_script"><![CDATA[
        data_input <- read.delim("$input_data", sep="\t")
        metadata <- read.delim("$input_annotations", sep="\t")
        runCol <- colnames(data_input)[$runcol]
        fcol_aggregation_pep <- colnames(data_input)[${peptide_aggregation.column_aggregation_peptides}]
        fcol_aggregation_prot <- colnames(data_input)[${protein_aggregation.column_aggregation_proteins}]

        dir.create("plots")

        scp <- scp::readSCP(
            assayData = data_input,
            colData = metadata,
            runCol = runCol,
            removeEmptyCols = $remove_empty_columns
        )
        number_of_assays <- length(scp)
        scp <- QFeatures::zeroIsNA(scp, i = 1:number_of_assays)

        #if $filtering_data.filter_reverse
        scp <- QFeatures::filterFeatures(scp,
                        ~ Reverse != "+")
        #end if

        #if $filtering_data.filter_contaminants
        scp <- QFeatures::filterFeatures(scp,
                        ~ Potential.contaminant != "+")
        #end if

        scp <- QFeatures::filterFeatures(scp,
                      ~ !is.na(PIF) & PIF > ${filtering_data.PIF_threshold})

        keepAssay <- QFeatures::dims(scp)[1, ] > ${filtering_data.minimum_features}
        scp <- scp[, , keepAssay]
        
        number_of_assays <- length(scp)
        single_cell_channels <- gsub(",", "|", "${filtering_data.single_cells}")

        scp <- scp::computeSCR(scp,
                  i = 1:number_of_assays,
                  colvar = "SampleType",
                  carrierPattern = "Carrier",
                  samplePattern = single_cell_channels,
                  sampleFUN = "mean",
                  rowDataName = "MeanSCR")
        
        #if $generate_QC_plots
        QC_plot_SCR <- QFeatures::rbindRowData(scp, i = 1:number_of_assays) |>
                        data.frame() |>
                        ggplot2::ggplot(ggplot2::aes(x = MeanSCR)) +
                        ggplot2::geom_histogram() +
                        ggplot2::geom_vline(xintercept = c(1/$count_cell_carrier, 0.1),
                                                lty = c(2, 1)) +
                        ggplot2::scale_x_log10()
        ggplot2::ggsave(filename = file.path("plots", "QC_plot_SCR.pdf"), QC_plot_SCR)
        
        QC_plot_SCR_col <- QFeatures::rbindRowData(scp, i = 1:number_of_assays) |>
                            data.frame() |>
                            ggplot2::ggplot(ggplot2::aes(x = MeanSCR, color = runCol)) +
                            ggplot2::geom_density() +
                            ggplot2::geom_vline(xintercept = 0.02, lty = 2) +
                            ggplot2::geom_vline(xintercept = 1, lty = 1)+
                            ggplot2::scale_x_log10()        
        ggplot2::ggsave(filename = file.path("plots", "QC_plot_SCR_col.pdf"), QC_plot_SCR_col)
        #end if

        scp <- QFeatures::filterFeatures(scp,
                      ~ !is.na(MeanSCR) &
                        MeanSCR < ${filtering_data.SCR_threshold})    

        #if $filtering_data.qvalue_level == "PSM"
        scp <- scp::pep2qvalue(scp,
                    i = 1:number_of_assays,
                    PEP = "dart_PEP",
                    rowDataName = "qvalue")
  
        scp <- QFeatures::filterFeatures(scp,
                        ~ qvalue < ${filtering_data.qvalue_threshold})
        #else
        scp <- scp::pep2qvalue(scp,
                    i = 1:number_of_assays,
                    PEP = "dart_PEP",
                    groupBy = "$filtering_data.qvalue_level",
                    rowDataName = "qvalue")
 
        scp <- QFeatures::filterFeatures(scp,
                        ~ qvalue < ${filtering_data.qvalue_threshold})
        #end if

        #if $filtering_data.divide_reference
        scp <- scp::divideByReference(scp,
                           i = 1:number_of_assays,
                           colvar = "SampleType",
                           samplePattern = ".",
                           refPattern = "Reference")
        #end if

        scp <- scp::aggregateFeaturesOverAssays(
            scp,
            i = 1:number_of_assays,
            fcol = fcol_aggregation_pep,
            name = paste0("peptide_", names(scp)),
            fun = ${peptide_aggregation.aggregation_peptides},
            na.rm = TRUE
        )

        scp <- QFeatures::joinAssays(scp,
                  i = grep("peptide", names(scp)),
                  name = "peptides")
   
        keep_samples <- unlist(strsplit("${peptide_filtering.samples_to_keep}", split=","))
        scp <- scp[,SummarizedExperiment::colData(scp)[["SampleType"]] %in% keep_samples, ]

        #if $peptide_filtering.filter_median_intensity.cut_median_intensity == "yes"
            medians <- colMedians(SummarizedExperiment::assay(scp[["peptides"]]), na.rm = TRUE)
            SummarizedExperiment::colData(scp)[["MedianRI"]] <- medians

            #if $generate_QC_plots
            QC_medianRI <- SummarizedExperimentcolData(scp) |>
                            data.frame() |>
                            ggplot2::ggplot() +
                            ggplot2::aes(x = MedianRI,
                                y = SampleType,
                                fill = SampleType) +
                            ggplot2::geom_boxplot() +
                            ggplot2::scale_x_log10()
            ggplot2::ggsave(filename = file.path("plots", "QC_medianRI.pdf"), QC_medianRI)
            #end if

            scp <- scp[, !is.na(SummarizedExperiment::colData(scp)[["MedianRI"]]) & SummarizedExperiment::colData(scp)[["MedianRI"]] < ${peptide_filtering.filter_median_intensity.median_intensity_threshold}, ]
        #end if

        #if $peptide_filtering.filter_median_CV.cut_median_CV == "yes"
        number_of_observations <- ${peptide_filtering.filter_median_CV.minimum_peptides_CV}
        CV_threshold <- ${peptide_filtering.filter_median_CV.median_CV_threshold}
        scp <- scp::medianCVperCell(scp,
                         i = 1:number_of_assays,
                         groupBy = "Leading.razor.protein",
                         nobs = number_of_observations,
                         norm = "div.median",
                         na.rm = TRUE,
                         colDataName = "MedianCV")

            #if $generate_QC_plots
            QC_medianCV <- MultiAssayExperiment::getWithColData(scp, "peptides") |>
                            SummarizedExperiment::colData() |>
                            data.frame() |>
                            ggplot2::ggplot(ggplot2::aes(x = MedianCV,
                                fill = SampleType)) +
                            ggplot2::geom_boxplot() +
                            ggplot2::geom_vline(xintercept = CV_threshold)
            ggplot2::ggsave(filename = file.path("plots", "QC_medianCV.pdf"), QC_medianCV)
            #end if

        scp <- scp[, !is.na(SummarizedExperiment::colData(scp)[["MedianCV"]]) & SummarizedExperiment::colData(scp)[["MedianCV"]] < CV_threshold, ]
        #end if
        
        #if $peptide_filtering.remove_blank
        scp <- scp[, SummarizedExperiment::colData(scp)[["SampleType"]] != "Blank", ]
        #end if

        #if $peptide_processing.normalization_method.choose_normalization == "simple"
        scp <- QFeatures::normalize(
            scp,
            i = "peptides",
            name = "peptides_norm",
            method = "${peptide_processing.normalization_method.normalize_simple_method}"
        )
        #else
        scp <- QFeatures::sweep(
            scp,
            i = "peptides", 
            MARGIN = 2,
            FUN = "/",
            STATS = ${peptide_processing.normalization_method.normalize_columns}(SummarizedExperiment::assay(scp[["peptides"]]), na.rm = TRUE),
            name = "peptides_norm_col"
        )

        scp <- QFeatures::sweep(
            scp,
            i = "peptides_norm_col",
            MARGIN = 1,
            FUN = "/",
            STATS = ${peptide_processing.normalization_method.normalize_rows}(SummarizedExperiment::assay(scp[["peptides_norm_col"]]),  na.rm = TRUE),
            name =  "peptides_norm"
        ) 
        #end if

        scp <- QFeatures::logTransform(
            scp,
            base = ${peptide_processing.base},
            i = "peptides_norm",
            name = "peptides_log"
        )

        #if $generate_QC_plots
        QC_boxplot_peptide <- create_boxplots(scp, "peptides", FALSE, "Peptides not normalized")
        QC_boxplot_peptide_norm <- create_boxplots(scp, "peptides_log", TRUE, "Peptides normalized")

        ggplot2::ggsave(filename = file.path("plots", "QC_boxplot_peptide.pdf"), QC_boxplot_peptide)
        ggplot2::ggsave(filename = file.path("plots", "QC_boxplot_peptide_norm.pdf"), QC_boxplot_peptide_norm)
        #end if

        #if $peptide_processing.remove_missing_peptides.remove_peptides == "yes"
        pNA <-  ${peptide_processing.remove_missing_peptides.pNA_peptides} / 100
        scp <- QFeatures::filterNA(scp, i = "peptides_log", pNA = pNA)
        #end if

        scp <- scp::aggregateFeaturesOverAssays(
            scp,
            i = "peptides_log",
            fcol = fcol_aggregation_prot,
            name = "proteins",
            fun = ${protein_aggregation.aggregation_proteins},
            na.rm = TRUE
        )

        #if $protein_processing.normalization_method_protein.choose_normalization_protein == "simple_prot"
        scp <- QFeatures::normalize(
            scp,
            i = "proteins",
            name = "proteins_norm",
            method = "${protein_processing.normalization_method_protein.normalize_simple_method_prot}"
        )
        #else
        scp <- QFeatures::sweep(
            scp,
            i = "proteins", 
            MARGIN = 2,
            FUN = "/",
            STATS = ${protein_processing.normalization_method_protein.normalize_columns_prot}(SummarizedExperiment::assay(scp[["proteins"]]), na.rm = TRUE),
            name = "proteins_norm_col"
        )

        scp <- QFeatures::sweep(
            scp,
            i = "proteins_norm_col",
            MARGIN = 1,
            FUN = "/",
            STATS = ${protein_processing.normalization_method_protein.normalize_rows_prot}(SummarizedExperiment::assay(scp[["proteins_norm_col"]]),  na.rm = TRUE),
            name =  "proteins_norm"
        ) 
        #end if

        #if $generate_QC_plots
        QC_boxplot_protein <- create_boxplots(scp, "proteins", TRUE, "Proteins not normalized")
        QC_boxplot_protein_norm <- create_boxplots(scp, "proteins_norm", TRUE, "Proteins normalized")
        
        ggplot2::ggsave(filename = file.path("plots", "QC_boxplot_protein.pdf"), QC_boxplot_protein)
        ggplot2::ggsave(filename = file.path("plots", "QC_boxplot_protein_norm.pdf"), QC_boxplot_protein_norm)
        
        pdf(file = file.path("plots", "QC_heatmap_proteins.pdf"))
        plot_heatmap(scp, "proteins_norm")
        dev.off()
        #end if

        scp <- QFeatures::impute(
            scp,
            i = "proteins_norm",
            name = "proteins_imptd",
            method = "knn",
            k = ${protein_processing.impute_k},
            rowmax = 1,
            colmax= 1,
            maxp = Inf,
            rng.seed = 1234
        )

        batch_colname <- colnames(metadata)[${batch_correction.select_batch_correction.batch_col}]
        #if $batch_correction.select_batch_correction.batch_correction_method == "combat"
        sce <- MultiAssayExperiment::getWithColData(scp, "proteins_imptd")
        batch <- SummarizedExperiment::colData(scp)[[batch_colname]]
        model <- stats::model.matrix(~ SampleType, data = SummarizedExperiment::colData(sce))
  
        SummarizedExperiment::assay(sce) <- sva::ComBat(
            dat = SummarizedExperiment::assay(sce),
            batch = batch,
            mod = model
        )

        scp <- QFeatures::addAssay(
            scp,
            y = sce,
            name = "proteins_batchC"
        )

        scp <- QFeatures::addAssayLinkOneToOne(
            scp,
            from = "proteins_imptd",
            to = "proteins_batchC"
        )
        #else
        sce <- MultiAssayExperiment::getWithColData(scp, "proteins_imptd")
        preserve_colname <- colnames(metadata)[${batch_correction.select_batch_correction.preserve_col}]

        SummarizedExperiment::assay(sce) <- limma::removeBatchEffect(
            SummarizedExperiment::assay(sce),
            group = sce[[preserve_colname]],
            batch = sce[[batch_colname]]
        )
        scp <- QFeatures::addAssay(scp,
                  y = sce,
                  name = "proteins_batchC")
        scp <- QFeatures::addAssayLinkOneToOne(scp,
                              from = "proteins_imptd",
                              to = "proteins_batchC")
        #end if 

        #if $dimensionality_reduction.PCA_computation.run_PCA == "yes"
        PCA_color <- colnames(metadata)[$dimensionality_reduction.PCA_computation.pca_coloring]
        scp[["proteins_batchC"]] <- scater::runPCA(
            scp[["proteins_batchC"]],
            ncomponents = ${dimensionality_reduction.PCA_computation.ncomponents_PCA},
            ntop = Inf,
            scale = TRUE,
            exprs_values = 1,
            name = "PCA"
        )
  
        pca <- scater::plotReducedDim(
            scp[["proteins_batchC"]],
            dimred = "PCA",
            colour_by = PCA_color,
            point_alpha = 1
        )

        ggplot2::ggsave(filename = file.path("plots", "PCA.pdf"), pca)

            #if $dimensionality_reduction.PCA_computation.UMAP_computation.run_UMAP == "yes"
            scp[["proteins_batchC"]] <- scater::runUMAP(
                scp[["proteins_batchC"]],
                ncomponents = ${dimensionality_reduction.PCA_computation.UMAP_computation.ncomponents_UMAP},
                ntop = Inf,
                scale = TRUE,
                exprs_values = 1,
                n_neighbors = 3,
                dimred = "PCA",
                name = "UMAP"
            )
  
            umap <- scater::plotReducedDim(
                scp[["proteins_batchC"]],
                dimred = "UMAP",
                colour_by = "SampleType",
                point_alpha = 1
            )
            ggplot2::ggsave(filename = file.path("plots", "UMAP.pdf"), umap)
            #end if
        #end if

        assay_df <- as.data.frame(SummarizedExperiment::assay(scp, "proteins_batchC"))
        row_metadata <- as.data.frame(SummarizedExperiment::rowData(scp[["proteins_batchC"]]))

        export_data <- cbind(row_metadata, as.data.frame(assay_df))
        write.table(export_data, file = '$Processed_data', sep = "\t", quote = F)

        #if $data_export.export_tables
        export_all_assays(scp)
        #end if

        #if $data_export.export_RData
        save(scp, file='$scp_object')
        #end if    
        ]]></configfile>
    </configfiles>
    <inputs>
        <expand macro="scp_param"/>
    </inputs>
    <outputs>
        <data name="Processed_data" format="tabular" label="Batch-corrected protein levels"/>
        <collection name="intermediate_outputs" type="list" label="Intermediate outputs" format="tabular">
            <discover_datasets pattern="__name_and_ext__" directory="outputs" />
            <filter>data_export['export_tables']</filter>
        </collection>
        <data name="scp_object" format="rds" label="scp object as .rds">
            <filter>data_export['export_RData']</filter>
        </data>
        <data name="script" format="txt" label="R script">
            <filter>data_export['export_R_script']</filter>
        </data>
        <collection name="plots" type="list" label="Plots">
            <discover_datasets pattern="__name_and_ext__" directory="plots" />
                <filter>generate_QC_plots or dimensionality_reduction['PCA_computation']['run_PCA'] == 'yes' or dimensionality_reduction['PCA_computation']['run_PCA'] == 'yes'</filter>
        </collection>
    </outputs>
    <tests>
        <test expect_num_outputs='2'>
            <param name="input_data" value="evidence_subset.txt"/>
            <param name="input_annotations" value="sampleAnnotation.txt"/>
            <param name="runcol" value="19"/>
            <param name="single_cells" value="Macrophage,Monocyte"/>
            <param name="samples_to_keep" value="Macrophage,Monocyte,Blank"/>
            <param name="batch_col" value="2"/>
            <param name="column_aggregation_peptides" value="6"/>
            <param name="column_aggregation_proteins" value="17"/>
            <param name="pca_coloring" value="4"/>
            <output name="Processed_data">
                <assert_contents>
                    <has_n_lines n="90"/>
                    <has_text text="E9PAV3"/>
                    <has_size size="625000" delta="20"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs='2'>
            <param name="input_data" value="evidence_subset.txt" />
            <param name="input_annotations" value="sampleAnnotation.txt"/>
            <param name="runcol" value="19"/>
            <param name="single_cells" value="Macrophage,Monocyte"/>
            <param name="samples_to_keep" value="Macrophage,Monocyte,Blank"/>
            <param name="batch_col" value="2"/>
            <param name="column_aggregation_peptides" value="6"/>
            <param name="column_aggregation_proteins" value="17"/>
            <param name="pca_coloring" value="4"/>
            <output_collection name="plots" type="list">
                <element name="PCA" file="PCA.pdf" ftype="pdf" compare="sim_size" delta="60"/>
                <element name="QC_boxplot_peptide" file="QC_boxplot_peptide.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_boxplot_peptide_norm" file="QC_boxplot_peptide_norm.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_boxplot_protein" file="QC_boxplot.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_boxplot_protein_norm" file="QC_boxplot_protein_norm.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_heatmap_proteins" file="QC_heatmap_proteins.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_medianCV" file="QC_medianCV.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_plot_SCR" file="QC_plot_SCR.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="QC_plot_SCR_col" file="QC_plot_SCR_col.pdf" ftype="pdf" compare="sim_size" delta="40"/>
                <element name="UMAP" file="UMAP.pdf" ftype="pdf" compare="sim_size" delta="200"/>
            </output_collection>
            <output name="Processed_data">
                <assert_contents>
                    <has_size size="625000" delta="20"/>
                    <has_n_lines n="90"/>
                    <has_text text="E9PAV3"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs='3'>
            <param name="input_data" value="evidence_subset.txt" />
            <param name="input_annotations" value="sampleAnnotation.txt"/>
            <param name="runcol" value="19"/>
            <param name="single_cells" value="Macrophage,Monocyte"/>
            <param name="samples_to_keep" value="Macrophage,Monocyte,Blank"/>
            <param name="column_aggregation_peptides" value="6"/>
            <param name="column_aggregation_proteins" value="17"/>
            <param name="batch_col" value="2"/>
            <param name="export_tables" value="true"/>
            <param name="pca_coloring" value="4"/>
            <output_collection name="intermediate_outputs" type="list" count="6"/>
            <output name="Processed_data">
                <assert_contents>
                    <has_size size="625000" delta="20"/>
                    <has_n_lines n="90"/>
                    <has_text text="E9PAV3"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs='4'>
            <param name="input_data" value="evidence_subset.txt" />
            <param name="input_annotations" value="sampleAnnotation.txt"/>
            <param name="runcol" value="19"/>
            <param name="single_cells" value="Macrophage,Monocyte"/>
            <param name="samples_to_keep" value="Macrophage,Monocyte,Blank"/>
            <param name="batch_col" value="2"/>
            <param name="export_RData" value="TRUE"/>
            <param name="export_R_script" value="TRUE"/>
            <param name="column_aggregation_peptides" value="6"/>
            <param name="column_aggregation_proteins" value="17"/>
            <param name="pca_coloring" value="4"/>
            <output name="Processed_data">
                <assert_contents>
                    <has_size size="625000" delta="20"/>
                    <has_n_lines n="90"/>
                    <has_text text="E9PAV3"/>
                </assert_contents>
            </output>
            <output name="script">
                <assert_contents>
                    <has_n_lines n="271"/>
                    <has_text text='ggplot2::ggsave(filename = file.path("plots", "PCA.pdf"), pca)'/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
        @GENERAL_HELP@
    ]]></help>
    <expand macro="citations" />
</tool>