comparison minimap2.xml @ 29:cbe067a674a9 draft default tip

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/main/tools/minimap2 commit a02980a757cba9680a7863a3e67e9e10d77fc9f6

28:f774b65c34d0

<?xml version="1.0"?>
<tool id="minimap2" name="Map with minimap2" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="20.01">
    <description>A fast pairwise aligner for genomic and spliced nucleotide sequences</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="edam_ontology"/>

    #if $io_options.eqx:
        --eqx
    #end if
    -t \${GALAXY_SLOTS:-4}
    reference.fa
    #if $fastq_input.fastq_input_selector in ['single', 'paired_iv']:
        '$fastq_input.fastq_input1'
    #else if $fastq_input.fastq_input_selector == 'paired':
         '$fastq_input.fastq_input1' '$fastq_input.fastq_input2'
    #else if $fastq_input.fastq_input_selector == 'paired_collection':
         '$fastq_input.fastq_input1.forward' '$fastq_input.fastq_input1.reverse'
    #end if
    #if str($io_options.output_format) in ('BAM', 'CRAM'):
        -a | samtools view --no-PG -hT reference.fa

        </conditional>
        <!-- start unchanged copy from bwa-mem -->
        <conditional name="fastq_input">
            <param name="fastq_input_selector" type="select" label="Single or Paired-end reads" help="Select between paired and single end data">
                <option value="single">Single</option>
                <option value="paired">Paired</option>
                <option value="paired_collection">Paired Collection</option>
                <option value="paired_iv">Paired Interleaved</option>
            </param>
            <!-- below, preset options are only offered for single-end input
            because paired-end alignment in minimap2 is only enabled with -x sr
            (see https://github.com/lh3/minimap2/issues/190) -->
            <when value="single">

                    <option value="splice:hq">Long-read splice alignment for PacBio CCS reads (same as `splice` but with -C5 -O6,24 -B4) (splice:hq)</option>
                    <option value="sr">Short single-end reads without splicing (-k21 -w11 --sr -A2 -B8 -O12,32 -E2,1 -r50 -p.5 -N20 -f1000,5000 -n2 -m20 -s40 -g200 -2K50m --heap-sort=yes --secondary=no) (sr)</option>
                    <option value="self-homology">Construct a self-homology map - use same genome as query and reference (-DP -k19 -w19 -m200) (self-homology)</option>
                </param>
            </when>
            <when value="paired">
                <param name="fastq_input1" type="data" format="fastqsanger,fastqsanger.gz,fasta" label="Select first set of reads" help="Specify dataset with forward reads"/>
                <param name="fastq_input2" type="data" format="fastqsanger,fastqsanger.gz,fasta" label="Select second set of reads" help="Specify dataset with reverse reads"/>
                <expand macro="pe_anaylsis_fixed_selector" />
            </when>
            <when value="paired_collection">
                <param name="fastq_input1" format="fastqsanger,fastqsanger.gz,fasta" type="data_collection" collection_type="paired" label="Select a paired collection" help="See help section for an explanation of dataset collections"/>
                <expand macro="pe_anaylsis_fixed_selector" />
            </when>
            <when value="paired_iv">
                <param name="fastq_input1" type="data" format="fastqsanger,fastqsanger.gz,fasta" label="Select fastq dataset" help="Specify dataset with interleaved reads"/>
                <expand macro="pe_anaylsis_fixed_selector" />
            </when>
        </conditional>
        <section name="indexing_options" title="Indexing options">
            <param argument="-H" type="boolean" optional="true" truevalue="-H" falsevalue="" label="Use homopolymer-compressed k-mer ?"/>

        </data>
    </outputs>
    <tests>
        <test expect_num_outputs="1">
            <!-- test single input -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test cram output -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <param name="output_format" value="CRAM"/>
            <output name="alignment_output" ftype="cram" file="minimap2-test1-fasta.cram" compare="sim_size" />
        </test>
        <test expect_num_outputs="1">
            <!-- test paired input -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="paired"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fastq1.fq"/>
            <param name="fastq_input2" ftype="fastqsanger" value="bwa-mem-fastq2.fq"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test1.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test paired input with one pair compressed -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="paired"/>
            <param name="fastq_input1" ftype="fastqsanger.gz" value="bwa-mem-fastq1.fq.gz"/>
            <param name="fastq_input2" ftype="fastqsanger" value="bwa-mem-fastq2.fq"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test1.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test collection input -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="paired_collection"/>
            <param name="fastq_input1">
                <collection type="paired">
                    <element name="forward" value="bwa-mem-fastq1.fq" />
                    <element name="reverse" value="bwa-mem-fastq2.fq" />
                </collection>
            </param>
            <output name="alignment_output" ftype="bam" file="minimap2-test2.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test data table reference -->
            <param name="reference_source_selector" value="cached" />
            <param name="ref_file" value="bwa-mem-mt-genome"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test alignment options -->
            <param name="reference_source_selector" value="cached" />
            <param name="min_occ_floor" value="1000"/>
            <param name="ref_file" value="bwa-mem-mt-genome"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <section name="alignment_options">
                <!-- the folowing settings correspond to the defaults for "sr"
                mode. The purpose is to check that all alignment params get
                parsed correctly. -->
                <param name="A" value="2" />

            </section>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>        
        <test expect_num_outputs="1">
            <!-- test paf output -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fastqsanger"  value="mini_reads.fq" />
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger"  value="mini_reads.fq" />
            <param name="analysis_type_selector" value="ava-ont"/>
            <param name="output_format" value="paf"/>
            <output name="alignment_output" ftype="paf" file="mini_reads.paf" />
        </test>
        <test expect_num_outputs="1">
            <!-- test self-homology mode -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="minimap2-self-homology.fasta" />
            <param name="fastq_input_selector" value="single" />
            <param name="fastq_input1" ftype="fasta" value="minimap2-self-homology.fasta" />
            <param name="analysis_type_selector" value="self-homology" />
            <output name="alignment_output" ftype="bam" file="minimap2-self-homology.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test mask-len option -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <section name="mapping_options">
                <param name="mask_len" value="100"/>
            </section>
            <output name="alignment_output" ftype="bam" file="minimap2-test-mask_len.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test map-hifi -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="pacbio_hifi_assembly.fa.gz"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="pacbio_hifi_reads.fasta.gz"/>
            <param name="analysis_type_selector" value="map-hifi"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test_hifi-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test map-hifi uncompressed reference-->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="pacbio_hifi_assembly.fa"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="pacbio_hifi_reads.fasta.gz"/>
            <param name="analysis_type_selector" value="map-hifi"/>
            <output name="alignment_output" ftype="bam" file="minimap2-test_hifi-2-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test kmer ocurrence interval option -->
            <param name="reference_source_selector" value="history" />
            <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            <param name="fastq_input_selector" value="single"/>
            <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
            <param name="analysis_type_selector" value="sr"/>
            <conditional name="kmer_ocurrence_interval">
                <param name="interval" value="enabled"/>
                <param name="lower_limit" value="10"/>
                <param name="upper_limit" value="30"/>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test-kmer_ocurrence.bam" lines_diff="4" />
        </test>
    </tests>
    <help>

Users’ Guide
------------

Minimap2 is a versatile sequence alignment program that aligns DNA or
mRNA sequences against a large reference database. Typical use cases
include: (1) mapping PacBio or Oxford Nanopore genomic reads to the

29:cbe067a674a9

<tool id="minimap2" name="Map with minimap2" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="@PROFILE@">
    <description>A fast pairwise aligner for genomic and spliced nucleotide sequences</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="edam_ontology"/>

    #if $io_options.eqx:
        --eqx
    #end if
    -t \${GALAXY_SLOTS:-4}
    reference.fa
    #if $fastq_input.fastq_input_selector == 'single':
        '$fastq_input.fastq_input1'
    #else if $fastq_input.fastq_input_selector == 'paired_collection':
         '$fastq_input.fastq_input1.forward' '$fastq_input.fastq_input1.reverse'
    #end if
    #if str($io_options.output_format) in ('BAM', 'CRAM'):
        -a | samtools view --no-PG -hT reference.fa

        </conditional>
        <!-- start unchanged copy from bwa-mem -->
        <conditional name="fastq_input">
            <param name="fastq_input_selector" type="select" label="Single or Paired-end reads" help="Select between paired and single end data">
                <option value="single">Single</option>
                <option value="paired_collection">Paired Collection</option>
            </param>
            <!-- below, preset options are only offered for single-end input
            because paired-end alignment in minimap2 is only enabled with -x sr
            (see https://github.com/lh3/minimap2/issues/190) -->
            <when value="single">

                    <option value="splice:hq">Long-read splice alignment for PacBio CCS reads (same as `splice` but with -C5 -O6,24 -B4) (splice:hq)</option>
                    <option value="sr">Short single-end reads without splicing (-k21 -w11 --sr -A2 -B8 -O12,32 -E2,1 -r50 -p.5 -N20 -f1000,5000 -n2 -m20 -s40 -g200 -2K50m --heap-sort=yes --secondary=no) (sr)</option>
                    <option value="self-homology">Construct a self-homology map - use same genome as query and reference (-DP -k19 -w19 -m200) (self-homology)</option>
                </param>
            </when>
            <when value="paired_collection">
                <param name="fastq_input1" format="fastqsanger,fastqsanger.gz,fasta" type="data_collection" collection_type="paired" label="Select a paired collection" help="See help section for an explanation of dataset collections"/>
                <expand macro="pe_anaylsis_fixed_selector" />
            </when>
        </conditional>
        <section name="indexing_options" title="Indexing options">
            <param argument="-H" type="boolean" optional="true" truevalue="-H" falsevalue="" label="Use homopolymer-compressed k-mer ?"/>

        </data>
    </outputs>
    <tests>
        <test expect_num_outputs="1">
            <!-- test single input -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test cram output -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <section name="io_options">
                <param name="output_format" value="CRAM"/>
            </section>
            <output name="alignment_output" ftype="cram" file="minimap2-test1-fasta.cram" compare="sim_size" />
        </test>
        <test expect_num_outputs="1">
            <!-- test paired input -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="paired_collection"/>
                <param name="fastq_input1">
                    <collection type="paired">
                        <element name="forward" value="bwa-mem-fastq1.fq" />
                        <element name="reverse" value="bwa-mem-fastq2.fq" />
                    </collection>
                </param>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test1.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test paired input with one pair compressed -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="paired_collection"/>
                <param name="fastq_input1">
                    <collection type="paired">
                        <element name="forward" value="bwa-mem-fastq1.fq.gz" />
                        <element name="reverse" value="bwa-mem-fastq2.fq" />
                    </collection>
                </param>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test1.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test data table reference -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="cached" />
                <param name="ref_file" value="bwa-mem-mt-genome"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test alignment options -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="cached" />
                <param name="ref_file" value="bwa-mem-mt-genome"/>
            </conditional>
            <section name="mapping_options">
                <param name="min_occ_floor" value="1000"/>
            </section>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <section name="alignment_options">
                <!-- the folowing settings correspond to the defaults for "sr"
                mode. The purpose is to check that all alignment params get
                parsed correctly. -->
                <param name="A" value="2" />

            </section>
            <output name="alignment_output" ftype="bam" file="minimap2-test1-fasta.bam" lines_diff="4" />
        </test>        
        <test expect_num_outputs="1">
            <!-- test paf output -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fastqsanger"  value="mini_reads.fq" />
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger"  value="mini_reads.fq" />
                <param name="analysis_type_selector" value="ava-ont"/>
            </conditional>
            <section name="io_options">
                <param name="output_format" value="paf"/>
            </section>
            <output name="alignment_output" ftype="paf" file="mini_reads.paf" />
        </test>
        <test expect_num_outputs="1">
            <!-- test self-homology mode -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="minimap2-self-homology.fasta" />
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single" />
                <param name="fastq_input1" ftype="fasta" value="minimap2-self-homology.fasta" />
                <param name="analysis_type_selector" value="self-homology" />
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-self-homology.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test mask-len option -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <section name="mapping_options">
                <param name="mask_len" value="100"/>
            </section>
            <output name="alignment_output" ftype="bam" file="minimap2-test-mask_len.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test map-hifi -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="pacbio_hifi_assembly.fa.gz"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="pacbio_hifi_reads.fasta.gz"/>
                <param name="analysis_type_selector" value="map-hifi"/>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test_hifi-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test map-hifi uncompressed reference-->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="pacbio_hifi_assembly.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="pacbio_hifi_reads.fasta.gz"/>
                <param name="analysis_type_selector" value="map-hifi"/>
            </conditional>
            <output name="alignment_output" ftype="bam" file="minimap2-test_hifi-2-fasta.bam" lines_diff="4" />
        </test>
        <test expect_num_outputs="1">
            <!-- test kmer ocurrence interval option -->
            <conditional name="reference_source">
                <param name="reference_source_selector" value="history" />
                <param name="ref_file" ftype="fasta" value="bwa-mem-mt-genome.fa"/>
            </conditional>
            <conditional name="fastq_input">
                <param name="fastq_input_selector" value="single"/>
                <param name="fastq_input1" ftype="fastqsanger" value="bwa-mem-fasta1.fa"/>
                <param name="analysis_type_selector" value="sr"/>
            </conditional>
            <section name="mapping_options">
                <conditional name="kmer_ocurrence_interval">
                    <param name="interval" value="enabled"/>
                    <param name="lower_limit" value="10"/>
                    <param name="upper_limit" value="30"/>
                </conditional>
            </section>
            <output name="alignment_output" ftype="bam" file="minimap2-test-kmer_ocurrence.bam" lines_diff="4" />
        </test>
    </tests>
    <help>

Users' Guide
------------

Minimap2 is a versatile sequence alignment program that aligns DNA or
mRNA sequences against a large reference database. Typical use cases
include: (1) mapping PacBio or Oxford Nanopore genomic reads to the