Fastq screen

modules/nf-core/fastqscreen/buildfromindex/environment.yml

@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "fastqscreen_buildfromindex"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::fastq-screen=0.15.3"

modules/nf-core/fastqscreen/buildfromindex/main.nf

@@ -0,0 +1,69 @@
+process FASTQSCREEN_BUILDFROMINDEX {
+    label 'process_single'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/fastq-screen:0.15.3--pl5321hdfd78af_0':
+        'biocontainers/fastq-screen:0.15.3--pl5321hdfd78af_0'}"
+
+    input:
+    val(genome_names)
+    path(indexes), stageAs: "dir*"
+
+    output:
+    path("FastQ_Screen_Genomes"), emit: database
+    path "versions.yml"         , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    dir = "FastQ_Screen_Genomes"
+    folder = indexes.collect { it.toString() }
+    database = [genome_names, folder].transpose()
+    copy_indexes = folder.collect { "cp -r ${it} $dir/${it}"}.join(" && ")
+
+    // Folder name and index (within folder) name could be different - use bash to look for index name
+    config = database
+        .collect { "########## ${it[0]} \nDATABASE ${it[0]} $dir/${it[1]}/${it[1] + '_to_be_replaced'}" }
+        .join("\n\n")
+        .replace("\n", "\\n")
+
+    """
+    mkdir $dir
+    $copy_indexes
+
+    echo "$config" > fastq_screen.conf
+    sed 's/\\\\n/\\n/g' fastq_screen.conf > $dir/fastq_screen.conf
+    echo "Replace folder name real index name"
+
+    for f in ${folder.join(' ')}
+    do
+        TO_REPLACE=\${f}_to_be_replaced
+
+        REPLACE_WITH=(\$f"/*.rev.1.bt2")
+        REPLACE_WITH=\$(basename \$REPLACE_WITH)
+        REPLACE_WITH=\${REPLACE_WITH%%.rev.1.bt2}
+
+        sed "s/\$TO_REPLACE/\$REPLACE_WITH/g" $dir/fastq_screen.conf > new_conf.conf
+        mv new_conf.conf $dir/fastq_screen.conf
+    done
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        fastqscreen: \$(echo \$(fastq_screen --version 2>&1) | sed 's/^.*FastQ Screen v//; s/ .*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    dir = "FastQ_Screen_Genomes"
+    """
+    mkdir $dir
+    touch $dir/fastq_screen.conf
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        fastqscreen: \$(echo \$(fastq_screen --version 2>&1) | sed 's/^.*FastQ Screen v//; s/ .*\$//')
+    END_VERSIONS
+    """
+}

modules/nf-core/fastqscreen/buildfromindex/meta.yml

@@ -0,0 +1,34 @@
+name: "fastqscreen_buildfromindex"
+description: Build fastq screen config file from bowtie index files
+keywords:
+  - build
+  - index
+  - genome
+  - reference
+tools:
+  - "fastqscreen":
+      description: "FastQ Screen allows you to screen a library of sequences in FastQ format against a set of sequence databases so you can see if the composition of the library matches with what you expect."
+      homepage: "https://www.bioinformatics.babraham.ac.uk/projects/fastq_screen/"
+      documentation: "https://stevenwingett.github.io/FastQ-Screen/"
+      tool_dev_url: "https://github.com/StevenWingett/FastQ-Screen/archive/refs/tags/v0.15.3.zip"
+      doi: "10.5281/zenodo.5838377"
+      licence: ["GPL-3.0-or-later"]
+input:
+  - genome_names:
+      type: string
+      description: List of names for each index
+  - indexes:
+      type: directory
+      description: Bowtie2 genome directories containing index files
+output:
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - database:
+      type: directory
+      description: fastq screen database folder containing config file and index folders
+      pattern: "FastQ_Screen_Genomes"
+authors:
+  - "@snesic"
+  - "@JPejovicApis"

modules/nf-core/fastqscreen/buildfromindex/tests/main.nf.test

@@ -0,0 +1,71 @@
+nextflow_process {
+
+    name "Test Process FASTQSCREEN_BUILDFROMINDEX"
+    script "../main.nf"
+    process "FASTQSCREEN_BUILDFROMINDEX"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "fastqscreen"
+    tag "fastqscreen/buildfromindex"
+    tag "bowtie2/build"
+
+    setup {
+
+        run("BOWTIE2_BUILD") {
+            script "../../../bowtie2/build/main.nf"
+            process {
+                """
+                input[0] =  Channel.from([
+                                    [[id: "sarscov2"], file(params.test_data['sarscov2']['genome']['genome_fasta'], checkIfExists: true)],
+                                    [[id: "human"]   , file(params.test_data['homo_sapiens']['genome']['genome_21_fasta'], checkIfExists: true)]
+                            ])
+                """
+            }
+        }
+
+    }
+
+    test("sarscov2 - human") {
+
+        when {
+            process {
+                """
+                input[0] = BOWTIE2_BUILD.out.index.map{meta, index -> meta.id}.collect()
+                input[1] = BOWTIE2_BUILD.out.index.map{meta, index -> index}.collect()
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.database, process.out.versions).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - human - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = BOWTIE2_BUILD.out.index.map{meta, index -> meta.id}.collect()
+                input[1] = BOWTIE2_BUILD.out.index.map{meta, index -> index}.collect()
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}

modules/nf-core/fastqscreen/buildfromindex/tests/main.nf.test.snap

@@ -0,0 +1,71 @@
+{
+    "database": {
+        "content": [
+            [
+                [
+                    [
+                        "genome.1.bt2:md5,cbe3d0bbea55bc57c99b4bfa25b5fbdf",
+                        "genome.2.bt2:md5,47b153cd1319abc88dda532462651fcf",
+                        "genome.3.bt2:md5,4ed93abba181d8dfab2e303e33114777",
+                        "genome.4.bt2:md5,c25be5f8b0378abf7a58c8a880b87626",
+                        "genome.rev.1.bt2:md5,52be6950579598a990570fbcf5372184",
+                        "genome.rev.2.bt2:md5,e3b4ef343dea4dd571642010a7d09597"
+                    ],
+                    [
+                        "genome.1.bt2:md5,2fbc8eeaf480f03b3a2362a782fa5755",
+                        "genome.2.bt2:md5,a5300e3cf590d4cd8bd5521cac6337d6",
+                        "genome.3.bt2:md5,d2a7a0e1c9891e847a15781c0695b84e",
+                        "genome.4.bt2:md5,a55dc4b09e3e00586de6963800c95b9a",
+                        "genome.rev.1.bt2:md5,93c07f7d04e5da6bf1322ad5f46ac806",
+                        "genome.rev.2.bt2:md5,c3fa30132c0c9b9bc7868059e8e23466"
+                    ],
+                    "fastq_screen.conf:md5,ca56e866c7006a46bdfe4e751e787265"
+                ]
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-05-01T20:37:34.647221"
+    },
+    "versions": {
+        "content": [
+            [
+                "versions.yml:md5,c2bcf85b00046c72d5dccd2dba8ac35c"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-05-01T20:37:35.126085"
+    },
+    "sarscov2 - human - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        "fastq_screen.conf:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,c2bcf85b00046c72d5dccd2dba8ac35c"
+                ],
+                "database": [
+                    [
+                        "fastq_screen.conf:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,c2bcf85b00046c72d5dccd2dba8ac35c"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "23.10.1"
+        },
+        "timestamp": "2024-05-01T20:39:14.913696"
+    }
+}

modules/nf-core/fastqscreen/buildfromindex/tests/tags.yml

@@ -0,0 +1,2 @@
+fastqscreen/buildfromindex:
+  - "modules/nf-core/fastqscreen/buildfromindex/**"

modules/nf-core/fastqscreen/fastqscreen/environment.yml

@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "fastqscreen_fastqscreen"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::fastq-screen=0.15.3"

modules/nf-core/fastqscreen/fastqscreen/main.nf

@@ -0,0 +1,54 @@
+process FASTQSCREEN_FASTQSCREEN {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/fastq-screen:0.15.3--pl5321hdfd78af_0':
+        'biocontainers/fastq-screen:0.15.3--pl5321hdfd78af_0'}"
+
+    input:
+    tuple val(meta), path(reads)  // .fastq files
+    path database
+
+    output:
+    tuple val(meta), path("*.txt") , emit: txt
+    tuple val(meta), path("*.png") , emit: png
+    tuple val(meta), path("*.html"), emit: html
+    path "versions.yml"            , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def args = task.ext.args ?: ""
+
+    """
+    fastq_screen --threads ${task.cpus} \\
+        --aligner bowtie2 \\
+        --conf ${database}/fastq_screen.conf \\
+        $reads \\
+        $args \\
+        --outdir .
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        fastqscreen: \$(echo \$(fastq_screen --version 2>&1) | sed 's/^.*FastQ Screen v//; s/ .*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch test_1_screen.html
+    touch test_1_screen.png
+    touch test_1_screen.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        fastqscreen: \$(echo \$(fastq_screen --version 2>&1) | sed 's/^.*FastQ Screen v//; s/ .*\$//')
+    END_VERSIONS
+    """
+
+}

modules/nf-core/fastqscreen/fastqscreen/meta.yml

@@ -0,0 +1,44 @@
+name: fastqscreen_fastqscreen
+description: Align reads to multiple reference genomes using fastq-screen
+keywords:
+  - align
+  - map
+  - fasta
+  - fastq
+  - genome
+  - reference
+tools:
+  - "fastqscreen":
+      description: "FastQ Screen allows you to screen a library of sequences in FastQ format against a set of sequence databases so you can see if the composition of the library matches with what you expect."
+      homepage: "https://www.bioinformatics.babraham.ac.uk/projects/fastq_screen/"
+      documentation: "https://stevenwingett.github.io/FastQ-Screen/"
+      tool_dev_url: "https://github.com/StevenWingett/FastQ-Screen/archive/refs/tags/v0.15.3.zip"
+      doi: "10.5281/zenodo.5838377"
+      licence: ["GPL-3.0-or-later"]
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - reads:
+      type: file
+      description: |
+        List of input FastQ files of size 1 and 2 for single-end and paired-end data,
+        respectively.
+  - database:
+      type: directory
+      description: fastq screen database folder containing config file and index folders
+      pattern: "FastQ_Screen_Genomes"
+output:
+  - fastq_screen:
+      type: directory
+      description: Output fastq_screen file containing alignment statistics
+      pattern: "*.{_fq_screen}"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@snesic"
+  - "@JPejovicApis"