Adding TRUST4 module

modules/nf-core/trust4/environment.yml

@@ -0,0 +1,9 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/environment-schema.json
+name: "trust4"
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - "bioconda::trust4=1.0.13"

modules/nf-core/trust4/main.nf

@@ -0,0 +1,74 @@
+process TRUST4 {
+    tag "$meta.id"
+    label 'process_medium'
+
+    conda "bioconda::trust4=1.0.13"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/trust4:1.0.13--h43eeafb_0':
+        'biocontainers/trust4:1.0.13--h43eeafb_0' }"
+
+    input:
+    tuple val(meta), path(bam), path(reads)
+    tuple val(meta2), path(fasta)
+    tuple val(meta3), path(vdj_reference)
+
+    output:
+    tuple val(meta), path("*.tsv")          , emit: tsv
+    tuple val(meta), path("*_airr.tsv")     , emit: airr_tsv
+    tuple val(meta), path("*_report.tsv")   , emit: report_tsv
+    tuple val(meta), path("*.fa")           , emit: fasta
+    tuple val(meta), path("*.out")          , emit: out
+    tuple val(meta), path("*.fq")           , emit: fq
+    path "versions.yml"                     , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    def bam_mode = bam ? "-b ${bam}" : ''
+    def single_end_mode = reads && meta.single_end ? "-u ${reads}" : ''
+    // reference is optional for fastq input
+    def reference = vdj_reference ? "--ref ${vdj_reference}" : ""
+    // separate forward from reverse pairs
+    def (forward, reverse) = reads.collate(2).transpose()
+    def paired_end_mode = reads && (meta.single_end == false) ? "-1 ${forward[0]} -2 ${reverse[0]}" : ''
+    """
+    echo $reference
+    run-trust4 \\
+        ${bam_mode} \\
+        ${single_end_mode} \\
+        ${paired_end_mode} \\
+        -t $task.cpus \\
+        -f ${fasta} \\
+        -o ${prefix} \\
+        ${reference} \\
+        $args
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        trust4: \$(run-trust4 2>&1 | grep -o 'v[0-9.]*-r[0-9]*' | sed 's/^/TRUST4 using /' )
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}_airr.tsv
+    touch ${prefix}_airr_align.tsv
+    touch ${prefix}_report.tsv
+    touch ${prefix}_assembled_reads.fa
+    touch ${prefix}_annot.fa
+    touch ${prefix}_cdr3.out
+    touch ${prefix}_raw.out
+    touch ${prefix}_final.out
+    touch ${prefix}_toassemble.fq
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        trust4: \$(run-trust4 2>&1 | grep -o 'v[0-9.]*-r[0-9]*' | sed 's/^/TRUST4 using /' )
+    END_VERSIONS
+    """
+}

modules/nf-core/trust4/meta.yml

@@ -0,0 +1,75 @@
+---
+# yaml-language-server: $schema=https://raw.githubusercontent.com/nf-core/modules/master/modules/meta-schema.json
+name: "trust4"
+description: Run TRUST4 on RNA-seq data
+keywords:
+  - sort
+  - example
+  - genomics
+tools:
+  - "trust4":
+      description: "TCR and BCR assembly from bulk or single-cell RNA-seq data"
+      homepage: "https://github.com/liulab-dfci/TRUST4"
+      documentation: "https://github.com/liulab-dfci/TRUST4"
+      tool_dev_url: "https://github.com/liulab-dfci/TRUST4"
+      licence: ["GPL v3"]
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - bam:
+      type: file
+      description: BAM file from bulk or single-cell RNA-seq data
+      pattern: "*.bam"
+  - reads:
+      type: file
+      description: List of input FastQ files of size 1 and 2 for single-end and paired-end data, respectively
+  - fasta:
+      type: file
+      description: Path to the fasta file coordinate and sequence of V/D/J/C genes
+  - ref:
+      type: file
+      description: Path to detailed V/D/J/C gene reference file, such as from IMGT database.
+
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. `[ id:'sample1', single_end:false ]`
+  - tsv:
+      type: file
+      description: tsv files created by TRUST4
+      pattern: "*.tsv"
+  - airr_tsv:
+      type: file
+      description: TRUST4 results in AIRR format
+      pattern: "*_airr.tsv"
+  - report_tsv:
+      type: file
+      description: TRUST4 report in tsv format
+      pattern: "*_report.tsv"
+  - fasta:
+      type: file
+      description: Fasta files created by TRUST4
+      pattern: "*.fa"
+  - out:
+      type: file
+      description: Further report files
+      pattern: "*.out"
+  - fq:
+      type: file
+      description: Fastq files created by TRUST4
+      pattern: "*.fq"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+
+authors:
+  - "@mapo9, @Joaodemeirelles"
+maintainers:
+  - "@mapo9"

modules/nf-core/trust4/tests/main.nf.test

@@ -0,0 +1,117 @@
+nextflow_process {
+
+    name "Test Process TRUST4"
+    script "../main.nf"
+    process "TRUST4"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "trust4"
+
+    test("sarscov2 - single_end") {
+
+        when {
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'test', single_end:true ], // meta map
+                    [], // bam
+                    [file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true)] // reads
+                ])
+                input[1] = [[ id:'test'],[file(params.test_data['sarscov2']['genome']['transcriptome_fasta'], checkIfExists: true)]] // fasta
+                input[2] = [[ id:'test'],[]] // ref
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.airr_tsv,
+                                process.out.fq,
+                                process.out.fasta,
+                                process.out.report_tsv,
+                                process.out.versions,
+                                ).match() },
+                { assert process.out.out[0][1][0] ==~ ".*/test_cdr3.out" },
+                { assert process.out.out[0][1][1] ==~ ".*/test_final.out" },
+                { assert process.out.out[0][1][2] ==~ ".*/test_raw.out" },
+                { assert process.out.tsv[0][1][0] ==~ ".*/test_airr.tsv" },
+                { assert process.out.tsv[0][1][1] ==~ ".*/test_airr_align.tsv" },
+                { assert process.out.tsv[0][1][2] ==~ ".*/test_report.tsv" }
+            )
+        }
+
+    }
+
+    test("sarscov2 - paired_end") {
+
+        when {
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'test', single_end:false ], // meta map
+                    [], // bam
+                    [file(params.test_data['sarscov2']['illumina']['test2_1_fastq_gz'], checkIfExists: true),
+                    file(params.test_data['sarscov2']['illumina']['test2_2_fastq_gz'], checkIfExists: true)] // reads
+                ])
+                input[1] = [[ id:'test'],[file(params.test_data['sarscov2']['genome']['transcriptome_fasta'], checkIfExists: true)]] // fasta
+                input[2] = [[ id:'test'],[]] // ref
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.airr_tsv,
+                                process.out.fq,
+                                process.out.fasta,
+                                process.out.report_tsv,
+                                process.out.versions,
+                                ).match() },
+                { assert process.out.out[0][1][0] ==~ ".*/test_cdr3.out" },
+                { assert process.out.out[0][1][1] ==~ ".*/test_final.out" },
+                { assert process.out.out[0][1][2] ==~ ".*/test_raw.out" },
+                { assert process.out.tsv[0][1][0] ==~ ".*/test_airr.tsv" },
+                { assert process.out.tsv[0][1][1] ==~ ".*/test_airr_align.tsv" },
+                { assert process.out.tsv[0][1][2] ==~ ".*/test_report.tsv" }
+            )
+        }
+
+    }
+
+    test("sarscov2 - single_end - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = Channel.of([
+                    [ id:'test', single_end:true ], // meta map
+                    [],
+                    [file(params.test_data['sarscov2']['illumina']['test_1_fastq_gz'], checkIfExists: true)]
+                ])
+                input[1] = [[ id:'test'],[file(params.test_data['sarscov2']['genome']['transcriptome_fasta'], checkIfExists: true)]]
+                input[2] = [[ id:'test'],[]]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.versions).match("versions") },
+                { assert process.out.out[0][1][0] ==~ ".*/test_cdr3.out" },
+                { assert process.out.out[0][1][1] ==~ ".*/test_final.out" },
+                { assert process.out.out[0][1][2] ==~ ".*/test_raw.out" },
+                { assert process.out.tsv[0][1][0] ==~ ".*/test_airr.tsv" },
+                { assert process.out.tsv[0][1][1] ==~ ".*/test_airr_align.tsv" },
+                { assert process.out.tsv[0][1][2] ==~ ".*/test_report.tsv" }
+            )
+        }
+
+    }
+
+}