<tool id="scoary" name="Scoary" version="@TOOL_VERSION@+galaxy0" profile="20.01">
    <description>calculates the assocations between all genes in the accessory genome and the traits</description>
    <macros>
        <token name="@TOOL_VERSION@">1.6.16</token>
    </macros>
    <xrefs>
        <xref type="bio.tools">scoary</xref>
    </xrefs>
    <requirements>
        <requirement type="package" version="@TOOL_VERSION@">scoary</requirement>
    </requirements>
    <version_command>scoary --version</version_command>
    <command detect_errors="exit_code"><![CDATA[
scoary

###########
## Input ##
###########

-t '$input_traits'
-g '$input_genes'

#if $input_restricts:
	-r '$input_restricts'
#end if

########################
## Additional Options ##
########################

#if len($additional_options.series_pc) != 0
    -p #echo " ".join([ "'%s'" % $s.pvalue for $i, $s in enumerate($additional_options.series_pc) ])
    -c #echo " ".join([ "'%s'" % $s.correction for $i, $s in enumerate($additional_options.series_pc) ])
#end if

#if $additional_options.permute != 0:
    -e str($additional_options.permute)
#end if

#if $additional_options.maxhits != 0:
    -m str($additional_options.maxhits)
#end if

$additional_options.collapse
$output_options.upgma

#if $input_newicktree:
    -n '$input_newicktree'
#end if

#########
## END ##
#########

--no-time
&&
tail -n +1 *.csv | sed "s/\,/\\t/g" > scoary_output.tsv &&
mv *.nwk scoary_output.nwk

    ]]></command>
    <inputs>
        <param name="input_traits" argument="-t" type="data" format="csv" label="Trait table"/>
        <param name="input_genes" argument="-g" type="data" format="csv" label="Gene Presence/Absence table from ROARY (default output)"/>
        <param name="input_restricts" optional="true" argument="-r" type="data" format="tabular" label="Table to analyze a subset of strains" />

        <!-- Additional Options -->
        <section name="additional_options" title="Additional Options">
            <repeat name="series_pc" title="P-value cutoff(s) and Correction(s)">
                <param name="pvalue" argument="-p" type="float" min="0" max="1.0" value="0.05" label="P-value cutoff for one Trait" help="SCOARY will not report genes with higher p-values than this (Default=1.0=All). Provide a single value (applied to all) or exactly as many values as correction criteria and in corresponding order (e.g., 0.05 0.05 for two traits)."/>
                <param name="correction" argument="-c" type="select" label="P-value correction" help="Apply the p-value corrections to the p-value cutoffs you have entered (Default = Individual p-value)." >
                    <option value="I" selected="true">Individual (naive) p-value</option>
                    <option value="B">Bonferroni adjusted p-value</option>
                    <option value="BH">Benjamini-Hochberg adjusted p</option>
                    <option value="PW">Best (lowest) pairwise comparison</option>
                    <option value="EPW">Entire range of pairwise comparison p-values</option>
                    <option value="P">Empirical p-value from permutations</option>
                </param>
            </repeat>
            <param name="permute" argument="-e" type="integer" min="0" value="0" label="Permutations" help="Perform N number of permutations of the significant results post-analysis. (Default = 0 = None)" />
            <param name="maxhits" argument="-m" type="integer" min="0" value="0" label="Maximal number of hits to report" help="SCOARY will only report the top max hits results per trait. (Default = 0 = All)" />
            <param argument="--collapse" type="boolean" checked="false" truevalue="--collapse" falsevalue="" label="Collapse correlated genes" help="Collapse correlated genes (genes that have identical distribution patterns in the sample) into merged units. (Default=false)"/>
            <param name="input_newicktree" optional="true" argument="-n" type="data" format="newick" label="Supply a custom tree (Newick format) for phylogenetic analyses instead instead of calculating it internally." />
        </section>

        <!-- Output Options -->
        <section name="output_options" title="Output Options" expanded="true">
            <param name="upgma" argument="-u" type="boolean" checked="false" truevalue="-u" falsevalue="" label="UPGMA tree" help="Calculate UPGMA tree to a newick file."/>
        </section>
    </inputs>
    <outputs>
        <data name="out_tabular" format="tabular" from_work_dir="scoary_output.tsv" label="${tool.name} on ${on_string}: Table" />
        <data name="out_newick" format="newick" from_work_dir="scoary_output.nwk" label="${tool.name} on ${on_string}: Tree">
            <filter>(output_options['upgma'] is True)</filter>
        </data>
    </outputs>
    <tests>
        <test expect_num_outputs="2">
            <param name="input_traits" ftype="csv" value="Tetracycline_resistance.csv" />
            <param name="input_genes" ftype="csv" value="Gene_presence_absence.csv" />
            <param name="upgma" value="Yes" />
            <repeat name="series_pc">
                <param name="pvalue" value="0.05"/>
                <param name="correction" value="I"/>
            </repeat>
            <output name="out_tabular" file="scoary_output.tsv" ftype="tabular" sort="true">
                <assert_contents>
                    <has_n_lines n="573" />
                    <has_line line="==&gt; Bogus_trait.results.csv &lt;==" />
                    <has_line line="==&gt; Tetracycline_resistance.results.csv &lt;==" />
                </assert_contents>
            </output>
            <output name="out_newick" file="scoary_output.nwk" ftype="newick" />
        </test>
        <test expect_num_outputs="2">
            <param name="input_traits" ftype="csv" value="Tetracycline_resistance.csv" />
            <param name="input_genes" ftype="csv" value="Gene_presence_absence.csv" />
            <param name="upgma" value="Yes" />
            <repeat name="series_pc">
                <param name="pvalue" value="0.05"/>
                <param name="correction" value="I"/>
            </repeat>
            <repeat name="series_pc">
                <param name="pvalue" value="0.05"/>
                <param name="correction" value="EPW"/>
            </repeat>
            <output name="out_tabular" file="scoary_output_2.tsv" ftype="tabular" sort="true">
                <assert_contents>
                    <has_n_lines n="27" />
                    <has_line line="==&gt; Bogus_trait.results.csv &lt;==" />
                    <has_line line="==&gt; Tetracycline_resistance.results.csv &lt;==" />
                </assert_contents>
            </output>
            <output name="out_newick" file="scoary_output_2.nwk" ftype="newick" />
        </test>
    </tests>

    <help><![CDATA[

.. class:: infomark

**What it does**

-------------------

**Scoary**

Scoary is designed to take the csv file from Roary as well as a traits file created by the user and calculate the assocations between all genes in the accessory genome and the traits. It reports a list of genes sorted by strength of association per trait.

-------------------

**Inputs**

-------------------

Scoary requires two input files: csv file from Roary and a list of traits to test associations to.
Traits can be anything as long as you can classify it into binary categories (e.g. antibiotic resistance, group membership (yes/no), MIC value higher/lower than 16).
Make sure you your entires are separated by ','.
The traits file needs to be formatted in a specific way (please take a look into the (documentation)[https://github.com/AdmiralenOla/Scoary]).

You can also use as input the pan-genome as called from Jason Sahl's program LS-BSR (Large-Scale Blast Score Ratio).
The program includes a python script for converting LS-BSR output to the Roary/Scoary format.

Trait presence is indicated by 1, trait absence by 0.
Assumes strain names in the first column and trait names in the first row.

Input gene presence/absence table (comma-separated-values) from ROARY.
Strain names must be equal to those in the trait table.

-----------

**Outputs**

-----------

Scory outputs a single csv traits file. It uses comma "," as a delimiter.
The results consists of genes that were found to be associated with the trait, sorted according to significance.
By default, Scoary reports all genes with a naive p-value < 0.05.

You can find the description of the columns in the (documentation)[https://github.com/AdmiralenOla/Scoary].

--------------------

**More Information**

--------------------

See the excellent `Scoary documentation`_

.. _`Scoary documentation`: https://github.com/AdmiralenOla/Scoary


**P-value cutoff (-p)**: For Fishers, Bonferronis, and Benjamini-Hochbergs tests, SCOARY will not report genes with higher p-values than this.
For empirical p-values, this is treated as an alpha level instead.
I.e. 0.02 will filter all genes except the lower and upper percentile from this test.
Run with "-p 1.0" to report all genes. Accepts standard form (e.g. 1E-8).
Provide a single value (applied to all) or exactly as many values as correction criteria and in corresponding order (e.g., 0.05 0.1 0.05 0.02).

**Correction (-c)**: Apply the indicated filtration measure: I=Individual (naive) p-value, B=Bonferroni adjusted p-value, BH=Benjamini-Hochberg adjusted p, PW=Best (lowest) pairwise comparison, EPW=Entire range of pairwise comparison p-values, P=Empirical p-value from permutations.
You can enter as many correction criteria as you would like.
These will be associated with the p-value cutoffs you enter.
For example "-c I EPW -p 0.1 0.05" will apply the following cutoffs: Naive p-value must be lower than 0.1 AND the entire range of pairwise comparison values are below 0.05 for this gene.
Note that the empirical p-values should be interpreted at both tails.
Therefore, running "-c P -p 0.05" will apply an alpha of 0.05 to the empirical (permuted) p-values, i.e. it will filter everything except the upper and lower 2.5 percent of the distribution.

**Permute (-e)**: Perform N number of permutations of the significant results post-analysis.
Each permutation will do a label switching of the phenotype and a new p-value is calculated according to this new dataset.
After all N permutations are completed, the results are ordered in ascending order, and the percentile of the original result in the permuted p-value distribution is reported.

--------------------

**Galaxy Wrapper Development**

--------------------

Author: Florian Heyl

	]]></help>
    <citations>
	   <citation type="doi">10.1038/s41467-020-15171-6</citation>
    </citations>
</tool>