From 3ae8cbfad754dc45cae7fb42315f2873a7e9e417 Mon Sep 17 00:00:00 2001
From: CMGG ICT Team <ict@cmgg.be>
Date: Wed, 11 May 2022 13:32:56 +0200
Subject: [PATCH] bowtie2: remove code duplication

---
 modules/bowtie2/align/main.nf        | 112 +++++++++++++--------------
 tests/modules/bowtie2/align/test.yml |  88 ++++-----------------
 2 files changed, 65 insertions(+), 135 deletions(-)

diff --git a/modules/bowtie2/align/main.nf b/modules/bowtie2/align/main.nf
index c233f955..18209bdf 100644
--- a/modules/bowtie2/align/main.nf
+++ b/modules/bowtie2/align/main.nf
@@ -1,11 +1,11 @@
 process BOWTIE2_ALIGN {
     tag "$meta.id"
-    label 'process_high'
+    label "process_high"
 
-    conda (params.enable_conda ? 'bioconda::bowtie2=2.4.4 bioconda::samtools=1.15.1 conda-forge::pigz=2.6' : null)
-    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'https://depot.galaxyproject.org/singularity/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:1744f68fe955578c63054b55309e05b41c37a80d-0' :
-        'quay.io/biocontainers/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:1744f68fe955578c63054b55309e05b41c37a80d-0' }"
+    conda (params.enable_conda ? "bioconda::bowtie2=2.4.4 bioconda::samtools=1.15.1 conda-forge::pigz=2.6" : null)
+    container "${ workflow.containerEngine == "singularity" && !task.ext.singularity_pull_docker_container ?
+        "https://depot.galaxyproject.org/singularity/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:1744f68fe955578c63054b55309e05b41c37a80d-0" :
+        "quay.io/biocontainers/mulled-v2-ac74a7f02cebcfcc07d8e8d1d750af9c83b4d45a:1744f68fe955578c63054b55309e05b41c37a80d-0" }"
 
     input:
     tuple val(meta), path(reads)
@@ -13,69 +13,61 @@ process BOWTIE2_ALIGN {
     val   save_unaligned
 
     output:
-    tuple val(meta), path('*.bam')    , emit: bam
-    tuple val(meta), path('*.log')    , emit: log
-    tuple val(meta), path('*fastq.gz'), emit: fastq, optional:true
+    tuple val(meta), path("*.bam")    , emit: bam
+    tuple val(meta), path("*.log")    , emit: log
+    tuple val(meta), path("*fastq.gz"), emit: fastq, optional:true
     path  "versions.yml"              , emit: versions
 
     when:
     task.ext.when == null || task.ext.when
 
     script:
-    def args = task.ext.args ?: ''
-    def args2 = task.ext.args2 ?: ''
+    def args = task.ext.args ?: ""
+    def args2 = task.ext.args2 ?: ""
     def prefix = task.ext.prefix ?: "${meta.id}"
+
+    def unaligned = ""
+    def reads_args = ""
+
     if (meta.single_end) {
-        def unaligned = save_unaligned ? "--un-gz ${prefix}.unmapped.fastq.gz" : ''
-        """
-        INDEX=`find -L ./ -name "*.rev.1.bt2" | sed 's/.rev.1.bt2//'`
-        [ -z "\$INDEX" ] && INDEX=`find -L ./ -name "*.rev.1.bt2l" | sed 's/.rev.1.bt2l//'`
-        [ -z "\$INDEX" ] && echo "BT2 index files not found" 1>&2 && exit 1
-        bowtie2 \\
-            -x \$INDEX \\
-            -U $reads \\
-            --threads $task.cpus \\
-            $unaligned \\
-            $args \\
-            2> ${prefix}.bowtie2.log \\
-            | samtools view -@ $task.cpus $args2 -bhS -o ${prefix}.bam -
-
-        cat <<-END_VERSIONS > versions.yml
-        "${task.process}":
-            bowtie2: \$(echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//')
-            samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
-            pigz: \$( pigz --version 2>&1 | sed 's/pigz //g' )
-        END_VERSIONS
-        """
+        unaligned = save_unaligned ? "--un-gz ${prefix}.unmapped.fastq.gz" : ""
+        reads_args = "-U ${reads}"
     } else {
-        def unaligned = save_unaligned ? "--un-conc-gz ${prefix}.unmapped.fastq.gz" : ''
-        """
-        INDEX=`find -L ./ -name "*.rev.1.bt2" | sed 's/.rev.1.bt2//'`
-        [ -z "\$INDEX" ] && INDEX=`find -L ./ -name "*.rev.1.bt2l" | sed 's/.rev.1.bt2l//'`
-        [ -z "\$INDEX" ] && echo "BT2 index files not found" 1>&2 && exit 1
-        bowtie2 \\
-            -x \$INDEX \\
-            -1 ${reads[0]} \\
-            -2 ${reads[1]} \\
-            --threads $task.cpus \\
-            $unaligned \\
-            $args \\
-            2> ${prefix}.bowtie2.log \\
-            | samtools view -@ $task.cpus $args2 -bhS -o ${prefix}.bam -
-
-        if [ -f ${prefix}.unmapped.fastq.1.gz ]; then
-            mv ${prefix}.unmapped.fastq.1.gz ${prefix}.unmapped_1.fastq.gz
-        fi
-        if [ -f ${prefix}.unmapped.fastq.2.gz ]; then
-            mv ${prefix}.unmapped.fastq.2.gz ${prefix}.unmapped_2.fastq.gz
-        fi
-
-        cat <<-END_VERSIONS > versions.yml
-        "${task.process}":
-            bowtie2: \$(echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//')
-            samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
-            pigz: \$( pigz --version 2>&1 | sed 's/pigz //g' )
-        END_VERSIONS
-        """
+        unaligned = save_unaligned ? "--un-conc-gz ${prefix}.unmapped.fastq.gz" : ""
+        reads_args = "-1 ${reads[0]} -2 ${reads[1]}"
     }
+
+    def samtools = "samtools view -@ $task.cpus --fast --with-header ${args2} > ${prefix}.bam"
+
+
+    """
+    INDEX=`find -L ./ -name "*.rev.1.bt2" | sed "s/.rev.1.bt2//"`
+    [ -z "\$INDEX" ] && INDEX=`find -L ./ -name "*.rev.1.bt2l" | sed "s/.rev.1.bt2l//"`
+    [ -z "\$INDEX" ] && echo "BT2 index files not found" 1>&2 && exit 1
+
+    bowtie2 \\
+        -x \$INDEX \\
+        ${reads_args} \\
+        --threads $task.cpus \\
+        ${unaligned} \\
+        $args \\
+        2> ${prefix}.bowtie2.log \\
+        | ${samtools}
+
+    if [ -f ${prefix}.unmapped.fastq.1.gz ]; then
+        mv ${prefix}.unmapped.fastq.1.gz ${prefix}.unmapped_1.fastq.gz
+    fi
+
+    if [ -f ${prefix}.unmapped.fastq.2.gz ]; then
+        mv ${prefix}.unmapped.fastq.2.gz ${prefix}.unmapped_2.fastq.gz
+    fi
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        bowtie2: \$(echo \$(bowtie2 --version 2>&1) | sed 's/^.*bowtie2-align-s version //; s/ .*\$//')
+        samtools: \$(echo \$(samtools --version 2>&1) | sed 's/^.*samtools //; s/Using.*\$//')
+        pigz: \$( pigz --version 2>&1 | sed 's/pigz //g' )
+    END_VERSIONS
+    """
 }
+
diff --git a/tests/modules/bowtie2/align/test.yml b/tests/modules/bowtie2/align/test.yml
index ea9aa72a..b3cb8330 100644
--- a/tests/modules/bowtie2/align/test.yml
+++ b/tests/modules/bowtie2/align/test.yml
@@ -1,83 +1,21 @@
-- name: bowtie2 align single-end
-  command: nextflow run ./tests/modules/bowtie2/align -entry test_bowtie2_align_single_end -c ./tests/config/nextflow.config -c ./tests/modules/bowtie2/align/nextflow.config
+- name: bowtie2 align test_bowtie2_align_single_end
+  command: nextflow run tests/modules/bowtie2/align -entry test_bowtie2_align_single_end -c tests/config/nextflow.config
   tags:
-    - bowtie2
     - bowtie2/align
+    - bowtie2
   files:
-    - path: ./output/bowtie2/test.bam
-    - path: ./output/bowtie2/test.bowtie2.log
-    - path: ./output/bowtie2/bowtie2/genome.3.bt2
-      md5sum: 4ed93abba181d8dfab2e303e33114777
-    - path: ./output/bowtie2/bowtie2/genome.2.bt2
-      md5sum: 47b153cd1319abc88dda532462651fcf
-    - path: ./output/bowtie2/bowtie2/genome.1.bt2
-      md5sum: cbe3d0bbea55bc57c99b4bfa25b5fbdf
-    - path: ./output/bowtie2/bowtie2/genome.4.bt2
-      md5sum: c25be5f8b0378abf7a58c8a880b87626
-    - path: ./output/bowtie2/bowtie2/genome.rev.1.bt2
-      md5sum: 52be6950579598a990570fbcf5372184
-    - path: ./output/bowtie2/bowtie2/genome.rev.2.bt2
-      md5sum: e3b4ef343dea4dd571642010a7d09597
+    - path: output/bowtie2/test.bam
+    - path: output/bowtie2/test.bowtie2.log
+      md5sum: 7b8a9e61b7646da1089b041333c41a87
+    - path: output/bowtie2/versions.yml
 
-- name: bowtie2 align paired-end
-  command: nextflow run ./tests/modules/bowtie2/align -entry test_bowtie2_align_paired_end -c ./tests/config/nextflow.config -c ./tests/modules/bowtie2/align/nextflow.config
+- name: bowtie2 align test_bowtie2_align_paired_end
+  command: nextflow run tests/modules/bowtie2/align -entry test_bowtie2_align_paired_end -c tests/config/nextflow.config
   tags:
-    - bowtie2
     - bowtie2/align
-  files:
-    - path: ./output/bowtie2/test.bam
-    - path: ./output/bowtie2/test.bowtie2.log
-    - path: ./output/bowtie2/bowtie2/genome.3.bt2
-      md5sum: 4ed93abba181d8dfab2e303e33114777
-    - path: ./output/bowtie2/bowtie2/genome.2.bt2
-      md5sum: 47b153cd1319abc88dda532462651fcf
-    - path: ./output/bowtie2/bowtie2/genome.1.bt2
-      md5sum: cbe3d0bbea55bc57c99b4bfa25b5fbdf
-    - path: ./output/bowtie2/bowtie2/genome.4.bt2
-      md5sum: c25be5f8b0378abf7a58c8a880b87626
-    - path: ./output/bowtie2/bowtie2/genome.rev.1.bt2
-      md5sum: 52be6950579598a990570fbcf5372184
-    - path: ./output/bowtie2/bowtie2/genome.rev.2.bt2
-      md5sum: e3b4ef343dea4dd571642010a7d09597
-
-- name: bowtie2 align single-end large-index
-  command: nextflow run ./tests/modules/bowtie2/align -entry test_bowtie2_align_single_end -c ./tests/config/nextflow.config -c ./tests/modules/bowtie2/align/nextflow.config --force_large_index
-  tags:
     - bowtie2
-    - bowtie2/align
   files:
-    - path: ./output/bowtie2/test.bam
-    - path: ./output/bowtie2/test.bowtie2.log
-    - path: ./output/bowtie2/bowtie2/genome.3.bt2l
-      md5sum: 8952b3e0b1ce9a7a5916f2e147180853
-    - path: ./output/bowtie2/bowtie2/genome.2.bt2l
-      md5sum: 22c284084784a0720989595e0c9461fd
-    - path: ./output/bowtie2/bowtie2/genome.1.bt2l
-      md5sum: 07d811cd4e350d56267183d2ac7023a5
-    - path: ./output/bowtie2/bowtie2/genome.4.bt2l
-      md5sum: c25be5f8b0378abf7a58c8a880b87626
-    - path: ./output/bowtie2/bowtie2/genome.rev.1.bt2l
-      md5sum: fda48e35925fb24d1c0785f021981e25
-    - path: ./output/bowtie2/bowtie2/genome.rev.2.bt2l
-      md5sum: 802c26d32b970e1b105032b7ce7348b4
-
-- name: bowtie2 align paired-end large-index
-  command: nextflow run ./tests/modules/bowtie2/align -entry test_bowtie2_align_paired_end -c ./tests/config/nextflow.config -c ./tests/modules/bowtie2/align/nextflow.config --force_large_index
-  tags:
-    - bowtie2
-    - bowtie2/align
-  files:
-    - path: ./output/bowtie2/test.bam
-    - path: ./output/bowtie2/test.bowtie2.log
-    - path: ./output/bowtie2/bowtie2/genome.3.bt2l
-      md5sum: 8952b3e0b1ce9a7a5916f2e147180853
-    - path: ./output/bowtie2/bowtie2/genome.2.bt2l
-      md5sum: 22c284084784a0720989595e0c9461fd
-    - path: ./output/bowtie2/bowtie2/genome.1.bt2l
-      md5sum: 07d811cd4e350d56267183d2ac7023a5
-    - path: ./output/bowtie2/bowtie2/genome.4.bt2l
-      md5sum: c25be5f8b0378abf7a58c8a880b87626
-    - path: ./output/bowtie2/bowtie2/genome.rev.1.bt2l
-      md5sum: fda48e35925fb24d1c0785f021981e25
-    - path: ./output/bowtie2/bowtie2/genome.rev.2.bt2l
-      md5sum: 802c26d32b970e1b105032b7ce7348b4
+    - path: output/bowtie2/test.bam
+    - path: output/bowtie2/test.bowtie2.log
+      md5sum: bd89ce1b28c93bf822bae391ffcedd19
+    - path: output/bowtie2/versions.yml