<tool id="proteinortho" name="Proteinortho" version="@TOOL_VERSION@+galaxy@WRAPPER_VERSION@" profile="@PROFILE@">
    <description>detects orthologous proteins/genes within different species</description>
    <macros>
        <import>proteinortho_macros.xml</import>
        <xml name="test_outputs">
            <output name="proteinortho">
                <assert_contents>
                    <has_line_matching expression="# Species\tGenes\tAlg\.-Conn\.\t.*"/>
                    <has_line_matching expression="[0-9]+\t[0-9]+\t.*"/>
                    <has_line_matching expression=".*(C|C2|E|L|M)_[0-9]+.*"/>
                </assert_contents>
            </output>
            <output name="blastgraph">
                <assert_contents>
                    <has_line_matching expression="# file_a\tfile_b"/>
                    <has_line_matching expression="# a\tb\tevalue_ab\tbitscore_ab\tevalue_ba\tbitscore_ba"/>
                    <has_line_matching expression="# (C|C2|E|L|M)\.fasta\t(C|C2|E|L|M)\.fasta"/>
                    <has_line_matching expression=".*(C|C2|E|L|M)_[0-9]+\t(C|C2|E|L|M)_[0-9]+.*"/>
                </assert_contents>
            </output>
            <output name="proteinorthograph">
                <assert_contents>
                    <has_line_matching expression="# file_a\tfile_b"/>
                    <has_line_matching expression="# a\tb\tevalue_ab\tbitscore_ab\tevalue_ba\tbitscore_ba(\tsame_strand\tsimscore)?"/>
                    <has_line_matching expression="# (C|C2|E|L|M)\.fasta\t(C|C2|E|L|M)\.fasta"/>
                    <has_line_matching expression=".*(C|C2|E|L|M)_[0-9]+\t(C|C2|E|L|M)_[0-9]+.*"/>
                </assert_contents>
            </output>
        </xml>
    </macros>
    <expand macro="requirements"/>
    <expand macro="version_command"/>
    <command detect_errors="exit_code"><![CDATA[
        ## the following ln-action is necessary, since the file names are used by proteinortho (output contains filenames => species names)
        #import re
        #for $f in $input_files#
            ln -sf '$f' '${re.sub('[^\w\-_.]', '_', f.element_identifier)}' &&
        #end for
        #if $synteny.synteny_options == "specified":
            #for $f in $synteny.input_files_syn#
                ln -sf '$f' '${re.sub('[^\w\-_.]', '_', f.element_identifier)}' &&
            #end for#
        #end if
        proteinortho 
            --project=result
            --cpus="\${GALAXY_SLOTS:-4}"
            --ram="\${GALAXY_MEMORY_MB:-16000}"
            #if $more_options.selfblast:
                $more_options.selfblast
            #end if
            #if $more_options.singles:
                $more_options.singles
            #end if
            --p=$p
            --e=$evalue
            #if $more_options.cov:
                --cov=$more_options.cov
            #end if
            #if $more_options.sim:
                --sim=`LC_NUMERIC=C awk "BEGIN {printf \"%.2f\",$more_options.sim/100}"`
            #end if
            #if $more_options.identity:
                --cov=$more_options.identity
            #end if
            #if $more_options.isoform != "no":
                --isoform=$more_options.isoform
            #end if
            #if $synteny.synteny_options == "specified":
                --synteny
                --dups=$synteny.dups
                --cs=$synteny.cs
                --alpha=$synteny.alpha
            #end if
            #for $f in $input_files#
                ${re.sub('[^\w\-_.]', '_', f.element_identifier)}
            #end for#
            #if $synteny.synteny_options == "specified":    
                #for $f in $synteny.input_files_syn#
                    ${re.sub('[^\w\-_.]', '_', f.element_identifier)}
                #end for#
            #end if
            2> >(sed -E "s/.\[([0-9]{1,2}(;[0-9]{1,2})?)?[mGK]//g" 1>&2)
        #if $synteny.synteny_options == "specified":
            &&
            mv result.poff-graph result.proteinortho-graph &&
            mv result.poff.tsv result.proteinortho.tsv &&
            mv result.poff.html result.proteinortho.html ;
        #end if
    ]]></command>
    <inputs>
        <param name="input_files" format="fasta" type="data" multiple="true" min="2" label="Select the input fasta files (>2)" help="The input fasta files. At least 2 are needed!"/>    
        <param argument="--p" type="select" label="Similarity comparision algorithm" help="In the first step of proteinortho an all-versus-all reciprocal best hit graph is build from the input files (using this algorithm).">
            <option value="diamond" selected="true">diamond (aminoacid sequences)</option>
            <option value="autoblast">auto detect NCBI-BLAST (protein and nucleotide sequences)</option>
            <option value="blastp">NCBI-BLASTP+ (protein sequences)</option>
            <option value="blastn">NCBI-BLASTN+ (nucleotide sequences)</option>
            <option value="lastp">Last (aminoacid sequences)</option>
            <option value="lastn">Last (nucleotide sequences)</option>
            <option value="blatp">BLAT (aminoacid sequences)</option>
            <option value="blatn">BLAT (nucleotide sequences)</option>
        </param>
        <param argument="--evalue" type="float" value="0.001" min="0" label="E-value threshold of the blast algorithm" help="This is the main parameter for the generation of the reciprocal best hit graph. Larger values results in more false positives (connections between proteins)."/>
        <param argument="--conn" type="float" value="0.1" min="0." max="10." label="Minimal algebraic connectivity" help="This is the main parameter for the clustering step. Choose larger values then more splits are done, resulting in more and smaller clusters."/>
        <section name="more_options" title="Additional Options" expanded="False">
            <param argument="--cov" type="integer" value="50" min="0" max="100" label="Minimal coverage of best blast alignments in %"/>
            <param argument="--sim" type="integer" value="95" min="0" max="100" label="Minimal sequence similarity in %"/>
            <param argument="--identity" type="integer" value="25" min="0" max="100" label="Minimal percent identity of best blast hits in %"/>
            <param argument="--selfblast" type="boolean" checked="false" truevalue="--selfblast" falsevalue="" label="Apply selfblast, detects paralogs without orthologs "/>
            <param argument="--singles" type="boolean" checked="false" truevalue="--singles" falsevalue="" label="Report singleton genes without any hit "/>
            <param argument="--isoform" type="select" label="Use isoform information" help="The reciprocal best hit graph is build using isoform information (isoforms are treated equivalent). For ncbi : simply add the additional files to the input (file names need to match). For uniprot : the isoforms need to contain the word isoform and the corresponding identifier. For trinity simply use the trinity output format.">
                <option value="no" selected="true">Don't use isoform information</option>
                <option value="ncbi">ncbi style (..._additional.fasta)</option>
                <option value="uniprot">uniprot style (...isoform of...)</option>
                <option value="trinity">trinity style (...i4)</option>
            </param>
        </section>
        <conditional name="synteny">
            <param name="synteny_options" type="select" label="Activate synteny feature (POFF)" help="To enhance the prediction accuracy, the relative order of genes (synteny) can be used as additional feature for the discrimination of orthologs. For more details see doi:10.1371/journal.pone.0105015.">
                <option value="no" selected="true">no</option>
                <option value="specified">yes</option>
            </param>
            <when value="no"/>
            <when value="specified">
                <param argument="--dups" type="integer" value="0" min="0" max="100" label="Number of reiterations for adjacencies heuristic, to determine duplicated regions"/>
                <param argument="--cs" type="integer" value="3" min="0" max="100" label="Size of a maximum common substring (MCS) for adjacency matches"/>
                <param argument="--alpha" type="float" value="0.5" min="0." max="1." label="Minimal percent identity of best blast hits"/>
                <param name="input_files_syn" type="data" format="gff" multiple="true" min="2" label="Select the GFF3 files matching the input fasta files" help="The GFF3 files need matching names with the input fasta files. If you provide mybacteria123.faa or mybacteria123.fasta ... then you need to provide mybacteria123.gff here accoringly. The attributes column (#9) must contain the attribute Name=GENE IDENTIFIER where GENE IDENTIFIER corresponds to the respective (protein) identifier in the FASTA input. For example see https://gitlab.com/paulklemm_PHD/proteinortho/-/blob/master/test/C.gff"/> 
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="blastgraph" format="tabular" label="${tool.name} on ${on_string}: RBH graph" from_work_dir="result.blast-graph"/>
        <data name="proteinortho" format="tabular" label="${tool.name} on ${on_string}: orthology-groups" from_work_dir="result.proteinortho.tsv"/>
        <data name="proteinorthograph" format="tabular" label="${tool.name} on ${on_string}: orthology-pairs" from_work_dir="result.proteinortho-graph"/>
    </outputs>
    <tests>
        <test expect_num_outputs="3"> <!-- test normal -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=diamond"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- various parameter -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="evalue" value="1"/>
            <param name="conn" value="1"/>
            <param name="cov" value="42"/>
            <param name="sim" value="42"/>
            <param name="identity" value="42"/>
            <param name="selfblast" value="true"/>
            <param name="singles" value="true"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=diamond"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- synteny -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="input_files_syn" value="L.gff,C.gff,C2.gff,E.gff,M.gff"/>
            <param name="synteny_options" value="specified"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=diamond"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- blast -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="p" value="blastp"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=blastp"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- auto blast -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="p" value="autoblast"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=autoblast"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- last -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="p" value="lastp"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=lastp"/>
            </assert_command>
        </test>
        <test expect_num_outputs="3"> <!-- blat -->
            <param name="input_files" value="L.fasta,C.fasta,C2.fasta,E.fasta,M.fasta"/>
            <param name="p" value="blastp"/>
            <expand macro="test_outputs"/>
            <assert_command>
                <has_text text="--p=blastp"/>
            </assert_command>
        </test>
    </tests>
    <help><![CDATA[Proteinortho with POFF - An orthology detection tool

**What it does**

Proteinortho is a tool to detect orthologous proteins/genes within different species (at least 2). 

  | It compares similarities of given gene/protein sequences and clusters them to find significant groups.
  | The algorithm was designed to handle large-scale data and can be applied to hundreds of species at one.
  | Details can be found in (doi:10.1186/1471-2105-12-124).
  | To enhance the prediction accuracy, the relative order of genes (synteny) can be used as additional feature for the discrimination of orthologs. The corresponding extension, namely PoFF (details see doi:10.1371/journal.pone.0105015), is already build in Proteinortho. 

----

**Proteinortho in a nutshell**

----

* **(i) Build adaptive reciprocal best hit graph (RBH)**

      | Using the blast algorithm (diamond,blast,blat,...) all input sequences are compared against each other.
      | If two proteins find each other with respect to multiple criteria like minimal evalue, similarity compared to the best hit, ... then a edge is drawn between the two proteins.
      | The result of this step is outputted to RBH

* **(ii) Cluster the RBH**

      | Using two clustering algorithms, edges are removed that weakly connect two connected components to reduce false positive hits.
      | The resulting connected components are outputted in orthology-groups / -PAIRS 

----

**Proteinortho output files**

----

* **RBH**

      | The result of the (i) step, the reciprocal best hit graph. 
      | First a comment line announces 2 species (# ecoli.faa   human.faa), then each line corresponds to a reciprocal best hit between 2 proteins/genes of the announced species. The output format is shown below.
      | *seqidA*,*seqidB* = the 2 ids/names of the proteins involved 
      | *evalue_ab* = evalue with seqidA as query and seqidB as part of the database 
      | *bitscore_ab* = bitscore with seqidA as query ...
      | *evalue_ba* = evalue with seqidB as query ...
      | ...

.. csv-table::
    
    seqidA,seqidB,evalue_ab,bitscore_ab,evalue_ba,bitscore_ba    

----

* **orthology-groups**

      | The result of the (ii) step, the clustered reciprocal best hit graph or the orthology groups.
      | Every line corresponds to an orthology group of proteins/genes. 
      | The first 3 columns characterize general properties of that group: number of proteins, species and the algebraic connectivity. The higher the algebraic connectivity the more edges are there and the better the group is connected to itself in general. 
      | Then a column for each species follows containing the proteins of that species. If a species contributes with more than one protein to a group of orthologs, then they are ordered by connectivity.

.. csv-table::
    
    Species,Genes,Alg.-Conn.   

----

* **orthology-pairs**

      | The same as orthology-groups but every edge is printed one-by-one here. The output is formatted the same as the RBH graph:

.. csv-table::
    
    seqidA,seqidB,evalue_ab,bitscore_ab,evalue_ba,bitscore_ba   

----

**Proteinortho-Tools for downstream analysis**

* `proteinortho grab proteins` : find gene(s)/protein(s) in a given fasta file and retrieve their sequence(s). You can also use a orthology-groups file.
* `proteinortho summary` : Summaries the orthology-pairs/RBH files to determine how the species are connected to each other.

More information can be found on github https://gitlab.com/paulklemm_PHD/proteinortho
]]>
    </help>
    <expand macro="citations"/>
</tool>