<tool id="tiara" name="tiara" version="@TOOL_VERSION@+galaxy0" profile="21.05">
    <description>Deep-learning-based approach for identification of eukaryotic sequences in the metagenomic data </description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="biotools"/>
    <expand macro="requirements"/>
    <command detect_errors="exit_code"><![CDATA[ 
        
        tiara 
        -t \${GALAXY_SLOTS:-4} 
        -i '$input' 
        -o '$output'

        #if $taxonomy_filter
            --tf #for $tf in $taxonomy_filter
                $tf
            #end for
        #end if   
        #if $probabilities
            --pr '$probabilities'
        #end if
        #if $min_len
            -m '$min_len'
        #end if
        #if $cutoff_stage1
            -p $cutoff_stage1
            #if $cutoff_stage2
            $cutoff_stage2
            #end if
        #end if
        #if $advanced_options.advance.customize_kmer_length == 'customize'
            --k1 $advanced_options.advance.first_stage_kmer
            --k2 $advanced_options.advance.second_stage_kmer
        #end if

    ]]></command>
    <inputs>
        <param name="input" type="data" format="fasta" label="input fasta,fasta.gz file"/>
        <param name="taxonomy_filter" type="select" multiple="true" optional="true" label="Write sequences to fasta,fasta.gz files specified in the arguments to this option." help="all refers to all classes present in input fasta (to separate fasta files).">
            <option value="mit">mitochondria</option>
            <option value="pla">plastid</option>
            <option value="bac">bacteria</option>
            <option value="arc">archea</option>
            <option value="euk">eukarya</option>
            <option value="unk">unknown</option>
            <option value="pro">prokarya</option>
            <option value="all">all</option>
        </param>
        <param argument="probabilities" type="boolean" truevalue="--pr" falsevalue="" checked="false" label="Add probabilities of individual classes for each sequence."/>
        <param argument="min_len" type="integer" value="3000" min="1000" optional="true" label="Minimum length of a sequence. Default: 3000 bp." help="Specify the desired minimum length in base pairs.Default value is 3000 bp and we do not recommend classifying sequences shorter than 1000 bp. "/>
        <param argument="cutoff_stage1" type="float" value="0.65" min="0.5" max="1" optional="true" label="Probability threshold for the first stage." help="Probability threshold needed for classification in the first stage. Default: 0.65." />
        <param argument="cutoff_stage2" type="float" value="0.65" min="0.5" max="1" optional="true" label="Probability threshold for the second stage." help="Probability threshold needed for classification in the second stage. Default: 0.65." /> 
        <section name="advanced_options" title="k-mer" expanded="true">
            <conditional name="advance">
                <param argument="customize_kmer_length" type="select" label="Advanced options">
                    <option value="default_options">No, Use param defaults</option>
                    <option value="customize">Yes, See full parameter list</option>
                </param>
                <when value="customize">
                    <param argument="first_stage_kmer" type="select" label="Select k-mer length used in the first stage of classification (Default: 6).">
                        <option value="4">k-mer length 4</option>
                        <option value="5">k-mer length 5</option>
                        <option value="6" selected="True">default k-mer length</option>
                    </param>
                    <param argument="second_stage_kmer" type="select" label="k-mer length used in the second stage of classification (Default: 7).">
                        <option value="4">k-mer length 4</option>
                        <option value="5">k-mer length 5</option>
                        <option value="6">k-mer length 6</option>
                        <option value="7" selected="True">default k-mer length</option>
                    </param>
                </when>
                <when value="default_options">
                    <!-- Define actions or defaults for the default option if necessary -->
                </when>
            </conditional>
        </section>
    </inputs>
    <outputs>
        <data name="output" format="txt" label="${tool.name} on ${on_string}: sequence ID, classification results"/>
    </outputs>
    <tests>
        <test expect_num_outputs="1">
            <param name="input" value="plast_fr.fasta.gz"/>
            <param name="taxonomy_filter" value="pla"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*sequence_id*"/>
                    <has_n_lines n="11" delta="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="mitplas1.fasta"/>
            <param name="taxonomy_filter" value="pla,mit"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*sequence_id*"/>
                    <has_n_lines n="30" delta="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="sample_all.fasta"/>
            <param name="taxonomy_filter" value="all"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*sequence_id*"/>
                    <has_n_lines n="51" delta="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="sample_all.fasta"/>
            <param name="taxonomy_filter" value="euk,bac,arc,unk"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*sequence_id*"/>
                    <has_n_lines n="51" delta="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="eukarya_fr.fasta"/>
            <param name="taxonomy_filter" value="euk"/>
            <param name="min_len" value="5000"/>
            <param name="cutoff_stage1" value="0.65"/>
            <param name="cutoff_stage2" value="0.60"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*sequence_id*"/>
                    <has_n_lines n="11" delta="5"/>
                </assert_contents>
            </output>
        </test>
        <test expect_num_outputs="1">
            <param name="input" value="bacteria_fr.fasta"/>
            <param name="taxonomy_filter" value="bac"/>
            <param name="min_len" value="5000"/>
            <param name="cutoff_stage1" value="0.65"/>
            <param name="cutoff_stage2" value="0.60"/>
            <param name="probabilities" value="true"/>
            <output name="output" ftype="txt">
                <assert_contents>
                    <has_text_matching expression=".*bac*"/>
                    <has_n_lines n="11" delta="5"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
What it does
============
Tiara is a Deep-learning-based approach for identification of eukaryotic sequences in the metagenomic data powered by PyTorch.

How it works
============
The sequences are classified in two stages:

First Stage:
Input: Sequences are classified into classes: archaea, bacteria, prokarya, eukarya, organelle, and unknown.
Output: Classifications for each sequence into one of the above classes.

Second Stage:
Input: Sequences labeled as organelle from the first stage.
Output: Further classification into mitochondria, plastid, or unknown.

Required Inputs
===============
The primary input for Tiara is metagenomic sequence data that needs classification.

Generated Outputs
=================
The output will be the sequences categorized into specific classes as described above.

Additional Resources
====================
For a more comprehensive understanding of tiara and detailed usage instructions, please visit the tiara GitHub repository:
tiara GitHub Repository: [https://github.com/ibe-uw/tiara]
        
    ]]></help>
    <expand macro="citations"/>
</tool>