annotate peakachu.xml @ 0:bc594df2aba3 draft

planemo upload for repository https://github.com/tbischler/PEAKachu commit 21413560ba7ac260a92531c08899c4a342d1660d

<tool id="peakachu" name="PEAKachu" version="0.1.0.0">
    <description>Calls Peaks in CLIP data</description>
    <requirements>
        <requirement type="package" version="3.6">python</requirement>
        <requirement type="package" version="0.1.0">peakachu</requirement>
    </requirements>
    <version_command>
    <![CDATA[
        peakachu --version
    ]]>
    </version_command>
    <command detect_errors="aggressive">
    <![CDATA[
        #for $i, $clib in enumerate($controlLibs):
            #if $clib:
                ln -s -f '$clib' ${i}.c.bam &&
                ln -s -f '$clib.metadata.bam_index' ${i}.c.bam.bai &&
            #end if
        #end for
        #for $j, $elib in enumerate($experimentLibs):
            ln -s -f '$elib' ${j}.e.bam &&
            ln -s -f '$elib.metadata.bam_index' ${j}.e.bam.bai &&
        #end for

        pwd &&
        mkdir ./tmp_output && 
        peakachu 
        ${mode.mode_selector}
        --exp_libs 
        #for $i, $elib in enumerate($experimentLibs):
            '${i}.e.bam'
        #end for
        #for $i, $clib in enumerate($controlLibs):
            #if $clib and $i == 0:
                --ctr_libs 
            #end if
            #if $clib:
                '${i}.c.bam'
            #end if
        #end for
        $pairwise_replicates
        $paired_end
        --max_insert_size $max_insert_size
        --features '$features'
        --sub_features '$sub_features'
        --max_proc "\${GALAXY_SLOTS:-1}"
        --output_folder ./tmp_output
        #if str($mode.mode_selector) == 'adaptive':
            --min_cluster_expr_frac $mode.min_cluster_expr_frac
            --min_block_overlap $mode.min_block_overlap
            --min_max_block_expr $mode.min_max_block_expr
        #elif str($mode.mode_selector) == 'window':
            --window_size $mode.window_size
            --step_size $mode.step_size
            --stat_test $mode.stat_test
            --het_p_val_threshold $mode.het_p_val_threshold
            --rep_pair_p_val_threshold $mode.rep_pair_p_val_threshold
        #end if
       --norm_method $mode.norm_method.norm_method_selector
        #if str($mode.norm_method.norm_method_selector) == 'manual':
            --size_factors $size_factors
        #end if
        --mad_multiplier $mad_multiplier
        --fc_cutoff $fc_cutoff
        --padj_threshold $padj_threshold

        &&
        head -n 1 -q ./tmp_output/peak_tables/*.csv | head -n 1 > peaks.tsv &&
        tail -n +2 -q ./tmp_output/peak_tables/*.csv >> peaks.tsv &&
        mv peaks.tsv '$peak_tables' &&
        mv ./tmp_output/plots/Initial*.png '$MA_plot'

    ]]>
    </command>
    <inputs>
        <param name="experimentLibs" type="data" format="bam" label="Experiment Libraries" multiple="True"/>
        <param name="controlLibs" type="data" format="bam" label="Control Libraries" multiple="True" optional="True"/>
        <param argument="--pairwise_replicates" type="boolean" truevalue="--pairwise_replicates" falsevalue="" checked="False" label="Pairwise Replicates" /> 
        <param argument="--paired_end" type="boolean" truevalue="--paired_end" falsevalue="" checked="False" label="Paired End" /> 
        <param argument="--max_insert_size" type="integer" value="50" label="Maximum Insert Size"/>
        <!-- The gff feature is not implemented, because the function can easily be accomplished with featureCount or other tools
        <param name="gffs" type="data" format="gff" label="Annotation" optional="True" multiple="True"/-->
        <param argument="--features" type="text" label="Features">
            <sanitizer>
                <valid initial="default"/>
            </sanitizer>
        </param>
        <param argument="--sub_features" type="text" label="Sub-Features">
            <sanitizer>
                <valid initial="default"/>
            </sanitizer>
        </param>

        <conditional name="mode">
            <param name="mode_selector" type="select" label="Select Mode" help="These modes work differently.">
                <option value="adaptive" selected="True">Adaptive</option>
                <option value="window">Window</option>
                <!-- The following options are not implemented because they are vastly different and should be implemented as their own tool, if need be.
                <option value="coverage">Coverage</option>
                <option value="consensus_peak">Consensus Peak</option-->
            </param>
            <when value="adaptive">
                <param argument="--min_cluster_expr_frac" label="Minimum cluster Expression Fraction" help="Minimum fraction of a block in a cluster for further consideration." type="float" value="0.01"/>
                <param argument="--min_block_overlap" label="Minimum Block Overlap" help="Minimum fraction of the width of blocks for merging." type="float" value="0.5"/>
                <param argument="--min_max_block_expr" label="Minimum Block Expression" help="Minimum fraction of expression of blocks for merging." type="float" value="0.1"/>
                <conditional name="norm_method">
                    <param name="norm_method_selector" type="select" label="Normalisation Method.">
                        <option value="deseq" selected="True">DESeq2</option>
                        <option value="manual">Manual</option>
                        <option value="none">None</option>
                    </param>
                    <when value="deseq"/>
                    <when value="none"/>
                    <when value="manual">
                        <param argument="--size_factors" label="Size Factors" type="text" help="Size factors have to be seperated by SPACE">
                            <sanitizer>
                                <valid initial="default"/>
                            </sanitizer>
                        </param>
                    </when>
                </conditional>
            </when>
            <when value="window">
                <param argument="--window_size" label="Window Size" type="integer" value="25"/>
                <param argument="--step_size" label="Step Size" type="integer" value="5"/>
                <param name="stat_test" type="select" label="Statistical Test">
                    <option value="gtest" selected="True">gtest</option>
                    <option value="deseq">DESeq2</option>
                </param>
                <conditional name="norm_method">
                    <param name="norm_method_selector" type="select" label="Normalisation Method.">
                        <option value="tmm" selected="True">TMM</option>
                        <option value="deseq">DESeq2</option>
                        <option value="count">Count</option>
                        <option value="manual">Manual</option>
                        <option value="none">None</option>
                    </param>
                    <when value="deseq"/>
                    <when value="tmm"/>
                    <when value="count"/>
                    <when value="none"/>
                    <when value="manual">
                        <param argument="--size_factors" label="Size Factors" type="text" help="Size factors have to be seperated by SPACE">
                            <sanitizer>
                                <valid initial="default"/>
                            </sanitizer>
                        </param>
                    </when>
                </conditional>
                <param argument="--het_p_val_threshold" label="Heterogeneous p-value Threshold" type="float" value="0.01"/>
                <param argument="--rep_pair_p_val_threshold" label="Paired p-value Threshold" type="float" value="0.05"/>
            </when>
        </conditional>
        <param argument="--mad_multiplier" label="Mad Multiplier" type="float" value="2.0"/>
        <param argument="--fc_cutoff" label="Fold Change Threshold" type="float" value="2.0"/>
        <param argument="--padj_threshold" type="float" label="Adjusted p-value Threshold" value="0.05"/>  
    </inputs>
    <outputs>
        <data format="tabular" name="peak_tables" label="${tool.name} ${mode.mode_selector} on ${on_string}: peaks"/>
        <data format="png" name="MA_plot" label="${tool.name} ${mode.mode_selector} on ${on_string}: MA plot"/>
    </outputs>
    <tests>
        <test>
            <param name="experimentLibs" value="test1_+xl.bam"/>
            <param name="controlLibs" value="test1_-xl.bam"/>
            <output name="peak_tables" ftype="tabular" file="test1_peaks.tsv"/>
            <output name="MA_plot" ftype="png" file="test1_MA.png"/>
        </test>
    </tests>
    <help>
    <![CDATA[
**PEAKachu**

PEAKachu is a tool for the accurate mapping of RBP binding sites based on CLIP-seq and RIP-seq data.
PEAKachu uses signal and control libraries (ideally more than three each) to detect binding sites.
It implements two peak calling approaches

**adaptive**

The adaptive approach applies a three-step procedure to detect regions that are significantly enriched over controls.

- blockbuster is applied to the pooled libraries to combine similar sets of reads into blocks
- blocks are decomposed into peaks by iteratively applying a block merging heuristic
- peaks with significant enrichment of signal over control libraries are determined using DESeq2

**windowed**

The windowed approach subdivides the genome into overlapping regions.
After filtering of lowly expressed regions and library normalization (either using manual size factors, TMM, or DESeq2), this approach determines significantly enriched windows using eiterh DESeq2 or repeated G-tests of goodness-of-fit.

    ]]>
    </help>
    <citations>
    </citations>
</tool>