<tool id="scpred_predict_labels" name="Scpred predict" version="@TOOL_VERSION@+galaxy0">
    <description>Make cell type predictions using trained model.</description>
     <macros>
        <import>scpred_macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <command detect_errors="exit_code"><![CDATA[
        scpred_predict.R --input-object "${input_object}" --pred-data "${pred_data}" --normalise-data "${normalise_data}"       --recompute-alignment "${recompute_alignment}" --reference-scaling "${reference_scaling}" --output-path "${output_path}"  

        #if $normalisation_method
        --normalisation-method "${normalisation_method}"
        #end if
        #if $scale_factor
        --scale-factor "${scale_factor}"
        #end if
        #if $threshold_level
        --threshold-level "${threshold_level}"
        #end if
        #if $max_iter_harmony
        --max-iter-harmony "${max_iter_harmony}"
        #end if

        ]]></command>
    <inputs>

        <param type="data" name="input_object" label="Input Seurat Object" format="rdata" help="Input Seurat object in .rds format" />
        <param type="data" name="pred_data" label="Query Data Matrix" format="rdata" help="Path to the input prediction matrix in .rds format"/>
        <param type="boolean" checked="false" name="normalise_data" label="Normalise Data"  help="Should the predicted expression data be normalised? Default: False"/>
        <param type="text" name="normalisation_method" optional="true" value="LogNormalize" label="Normalisation Method" help="What normalisation method should be applied to predicted data? Default: LogNormalize"  />
        <param type="integer" name="scale_factor" optional="true" value="1000000" label="Scale Factor" help="What scale factor should be applied? Note: for CPM normalisation, 1e6 is selected by default"  />
        <param type="float" name="threshold_level" optional="true" value="0.8" label="Threshold Level" help="Classification threshold value"  />
        <param type="integer" name="max_iter_harmony" optional="true" value="20" label="Max Iterations" help="Maximum number of rounds to run Harmony. One round of Harmony involves one clustering and one correction step"  />
        <param type="boolean" checked="true" name="recompute_alignment" label="Recompute Alignment" help="Recompute alignment? Useful if scPredict() has already been run. Default: True"/>
        <param type="boolean" checked="true" name="reference_scaling" label="Reference Scaling" help="Scale new dataset based on means and stdevs from reference dataset before alignment. Default: True"/>

    </inputs>
    <outputs>
        <data name="output_path" format="rdata" />
    </outputs>
    <tests>
        <test>
            <param name="input_object" value="scPred_trained.rds" />
            <param name="normalise_data" value="True" />
            <param name="pred_data" value="query_pbmc.rds" />
            <output name="output_path" file="predicted_data.rds" compare="sim_size" delta="10000000"/>
        </test>
    </tests>
     <help><![CDATA[
    @HELP@
    
    @VERSION_HISTORY@
    ]]></help>
    <expand macro="citations" />
</tool>