<tool id="music_deconvolution" name="MuSiC" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@"
      profile="21.09" license="GPL-3.0-or-later" >
    <description>estimate cell type proportions in bulk RNA-seq data</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <command detect_errors="exit_code" ><![CDATA[
mkdir report_data &&
Rscript --vanilla '$__tool_directory__/scripts/${do.method}.R' '$conf'
]]></command>
    <configfiles>
        <configfile name="conf" >

null_str_vec = function(gstr){
   tokens = unlist(as.vector(strsplit(gstr, split=",")))
   if (length(tokens) == 0){
      return(NULL)
   }
   if (length(tokens) == 1){
      return(tokens[[1]])
   }
   return(tokens)
}

bulk_eset = readRDS('$bulk_eset')
scrna_eset = readRDS('$scrna_eset')

#if str($do.method) == "estimateprops":

phenotype_factors = null_str_vec('$do.phenotype_factors')
phenotype_factors_always_exclude = null_str_vec('$do.phenotype_factors_always_exclude')
celltypes_label = null_str_vec('$do.celltypes_label')
samples_label = null_str_vec('$do.samples_label')
celltypes = null_str_vec('$do.celltypes')
methods = c("MuSiC", "NNLS")
phenotype_target = null_str_vec('$do.phenotype_target')
phenotype_target_threshold = as.numeric('$do.phenotype_target_threshold')
sample_disease_group = null_str_vec('$do.sample_disease_group')
sample_disease_group_scale = as.integer('$do.sample_disease_group_scale')
compare_title = null_str_vec('$do.compare_title')

outfile_pdf='$out_pdf'

#elif str($do.method) == "dendrogram":

celltypes_label = null_str_vec('$do.celltypes_label')
clustertype_label = null_str_vec('$do.clustertype_label')
samples_label = null_str_vec('$do.samples_label')
celltypes = null_str_vec('$do.celltypes')

data.to.use = list(
    #for $i, $repeat in enumerate( $do.cluster_groups )
        #if $i == 0:
        $repeat.cluster_id = list(cell.types = null_str_vec('$repeat.celltypes'),
            marker.names = null_str_vec('$repeat.marker_name'),
                     marker.list = read_list('$repeat.marker_list'))
        #else
        , $repeat.cluster_id = list(cell.types = null_str_vec('$repeat.celltypes'),
                     marker.names = null_str_vec('$repeat.marker_name'),
                     marker.list = read_list('$repeat.marker_list'))
        #end if
    #end for
)

outfile_pdf='$out_pdf'
outfile_tab='$out_tab'

#else
    stop("No such option")
#end if

        </configfile>
    </configfiles>
    <inputs>
        <param name="scrna_eset" label="scRNA Dataset" type="data" format="rdata.eset" />
        <param name="bulk_eset" label="Bulk RNA Dataset" type="data" format="rdata.eset" />
        <conditional name="do" >
            <param name="method" type="select" label="Purpose" >
                <!-- The values here correspond to script names in the scripts folder
                     and must remain so -->
                <option value="estimateprops">Estimate Proportions</option>
                <option value="dendrogram">Compute Dendrogram</option>
            </param>
            <when value="estimateprops" >
                <param name="celltypes_label" type="text" value="cellType"
                       label="Cell Types Label from scRNA dataset" >
                    <expand macro="validator_text" />
                </param>
                <param name="samples_label" type="text" value="sampleID"
                       label="Samples Identifier from scRNA dataset" >
                    <expand macro="validator_text" />
                </param>
                <expand macro="celltypes_macro" />
                <param name="phenotype_factors" type="text"
                       label="Phenotype factors"
                       help="List of phenotypes factors to be used in the linear regression. Please make sure that each factor has more than one unique value. Names correspond to column names in the bulk RNA dataset phenotype table. If blank, then treat all bulk phenotype columns as factors." >
                    <expand macro="validator_index_identifiers" />
                </param>
                <param name="phenotype_factors_always_exclude" type="text"
                       label="Excluded phenotype factors"
                       help="List of phenotype factors to always exclude in the analysis"
                       value="sampleID,SubjectName" >
                    <expand macro="validator_index_identifiers" />
                </param>
                <param name="phenotype_target" type="text" label="Phenotype Target"
                       help="MUST exist in the bulk RNA datasets phenotype factors, as above." >
                    <expand macro="validator_text" />
                </param>
                <param name="phenotype_target_threshold" type="float" label="Phenotype Target Threshold"
                       value="-99"
                       help="The (%) threshold at which the phenotype target manifests. Leave at -99 to select all." >
                </param>
                <param name="sample_disease_group" type="text" label="Sample Disease Group"
                       help="MUST exist in the sample_groups above." >
                    <expand macro="validator_text" />
                </param>
                <param name="sample_disease_group_scale" type="integer"
                       label="Sample Disease Group (Scale)" value="5"
                       help="Used to accentutate certain features in the plots. Increase this number to reduce the effect." />
                <param name="compare_title" type="text" label="Plot Title" >
                    <expand macro="validator_text" />
                </param>
            </when>
            <when value="dendrogram" >
                <param name="celltypes_label" type="text" value="cellType"
                       label="Cell Types Label from scRNA dataset" >
                    <expand macro="validator_text" />
                </param>
                <param name="clustertype_label" type="text" value="clusterType"
                       label="Cluster Types Label from scRNA dataset" >
                    <expand macro="validator_text" />
                </param>
                <param name="samples_label" type="text" value="sampleID"
                       label="Samples Identifier from scRNA dataset" >
                    <expand macro="validator_text" />
                </param>
                <expand macro="celltypes_macro" />
                <repeat name="cluster_groups" title="Cluster Groups" min="0"
                        help="Insert cell cluster groups based on a previous clustering." >
                    <param name="cluster_id" label="Cluster ID" type="text" value=""
                           help="e.g. C1 or Cluster1, etc." />
                    <expand macro="celltypes_macro" />
                    <param name="marker_name" label="Marker Gene Group Name" type="text"
                           optional="true" value=""
                           help="Name of the list of gene markers used to describe the marker list supplied below." >
                        <expand macro="validator_text" />
                    </param>
                    <param name="marker_list" label="List of Gene Markers" type="data" format="txt,tabular"
                           optional="true"
                           help="A single column of marker genes" />
                </repeat>
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="out_pdf" format="pdf" label="${tool.name} on ${on_string}: PDF Plots" />
        <data name="out_tab" format="tabular" label="${tool.name} on ${on_string}: Cell Proportions by Sample" >
            <filter>do["method"] == "dendrogram" and len(do["cluster_groups"]) > 0</filter>
        </data>
        <collection name="props" type="list" label="${tool.name} on ${on_string}: Proportion Matrices" >
            <filter>do["method"] == "estimateprops"</filter>
            <discover_datasets pattern="prop_(?P&lt;designation&gt;.+)\.tabular" format="tabular" directory="report_data" />
        </collection>
        <collection name="summaries" type="list" label="${tool.name} on ${on_string}: Summaries and Logs">
            <filter>do["method"] == "estimateprops"</filter>
            <discover_datasets pattern="summ_(?P&lt;designation&gt;.+)\.txt" format="txt" directory="report_data" />
            <discover_datasets pattern="varprop_(?P&lt;designation&gt;.+)\.tabular" format="tabular" directory="report_data" />
            <discover_datasets pattern="rsquared_(?P&lt;designation&gt;.+)\.tabular" format="tabular" directory="report_data" />
            <discover_datasets pattern="weightgene_(?P&lt;designation&gt;.+)\.tabular" format="tabular" directory="report_data" />
        </collection>
    </outputs>
    <tests>
        <test expect_num_outputs="1" >
            <!-- Dendrogram test 1 -->
            <param name="bulk_eset" value="Mousebulkeset.rds" />
            <param name="scrna_eset" value="Mousesubeset.degenesonly2.half.rds" />
            <conditional name="do" >
                <param name="method" value="dendrogram" />
                <param name="celltypes_label" value="cellType" />
                <param name="samples_label" value="sampleID" />
                <param name="celltypes" value="Endo,Podo,PT,LOH,DCT,CD-PC,CD-IC,Fib,Macro,Neutro,B lymph,T lymph,NK" />
            </conditional>
            <output name="out_pdf" value="dendro_1.pdf" compare="sim_size" />
        </test>
        <test expect_num_outputs="2" >
            <!-- Dendrogram test 2 -->
            <param name="bulk_eset" value="Mousebulkeset.rds" />
            <param name="scrna_eset" value="Mousesubeset.degenesonly2.half.rds" />
            <conditional name="do" >
                <param name="method" value="dendrogram" />
                <param name="celltypes_label" value="cellType" />
                <param name="samples_label" value="sampleID" />
                <param name="celltypes" value="Endo,Podo,PT,LOH,DCT,CD-PC,CD-IC,Fib,Macro,Neutro,B lymph,T lymph,NK" />
                <repeat name="cluster_groups" >
                    <param name="cluster_id" value="C1" />
                    <param name="celltypes" value="Neutro" />
                </repeat>
                <repeat name="cluster_groups" >
                    <param name="cluster_id" value="C2" />
                    <param name="celltypes" value="Podo" />
                </repeat>
                <repeat name="cluster_groups" >
                    <param name="cluster_id" value="C3" />
                    <param name="celltypes" value="Endo,CD-PC,LOH,CD-IC,DCT,PT" />
                    <param name="marker_name" value="Epithelial" />
                    <param name="marker_list" value="epith.markers" />
                </repeat>
                <repeat name="cluster_groups" >
                    <param name="cluster_id" value="C4" />
                    <param name="celltypes" value="Macro,Fib,B lymph,NK,T lymph" />
                    <param name="marker_name" value="Immune" />
                    <param name="marker_list" value="immune.markers" />
                </repeat>
            </conditional>
            <output name="out_pdf" value="dendro.pdf" compare="sim_size" />
            <output name="out_tab">
                <assert_contents>
                    <has_text_matching expression="^\s+Neutro\s+Podo\s+Endo" />
                    <has_text text="APOL1.GNA78M"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="3" >
            <!-- Estimate Proportions test -->
            <param name="bulk_eset" value="GSE50244bulkeset.subset.rds" />
            <param name="scrna_eset" value="EMTABesethealthy.subset.rds" />
            <conditional name="do" >
                <param name="method" value="estimateprops" />
                <param name="celltypes_label" value="cellType" />
                <param name="samples_label" value="sampleID" />
                <param name="celltypes" value="alpha,beta,delta,gamma,acinar,ductal" />
                <param name="phenotype_factors" value="age,bmi,hba1c,gender" />
                <param name="phenotype_target" value="hba1c" />
                <param name="phenotype_target_threshold" value="6.5" />
                <param name="sample_disease_group" value="T2D" />
                <param name="sample_disease_group_scale" value="5" />
                <param name="compare_title" value="HbA1c vs Beta Cell Type Proportion" />
            </conditional>
            <output name="out_pdf" value="default_output.pdf" compare="sim_size" />
            <output_collection name="summaries" count="5">
                <element name="Log of MuSiC fitting" ftype="txt">
                    <assert_contents>
                        <has_text text="Residual standard error: 0.1704 on 72 degrees of freedom"/>
                    </assert_contents>
                </element>
                <element name="Log of NNLS fitting" ftype="txt">
                    <assert_contents>
                        <has_text text="Residual standard error: 0.0645 on 72 degrees of freedom"/>
                    </assert_contents>
                </element>
            </output_collection>
        </test>
    </tests>
    <help><![CDATA[
MuSiC utilizes cell-type specific gene expression from single-cell RNA sequencing (RNA-seq) data to characterize cell type compositions from bulk RNA-seq data in complex tissues. By appropriate weighting of genes showing cross-subject and cross-cell consistency, MuSiC enables the transfer of cell type-specific gene expression information from one dataset to another.

Solid tissues often contain closely related cell types which leads to collinearity. To deal with collinearity, MuSiC employs a tree-guided procedure that recursively zooms in on closely related cell types. Briefly, we first group similar cell types into the same cluster and estimate cluster proportions, then recursively repeat this procedure within each cluster.

.. image:: $PATH_TO_IMAGES/FigureMethod.jpg

    ]]></help>
    <citations>
        <citation type="doi">https://doi.org/10.1038/s41467-018-08023-x</citation>
    </citations>
</tool>