Update kalign module to work with compressed files. (#5277)

* add compression handling to kalign

* push conda to version 3.4

* fix version number

* commit tests

* update snapshot

---------

Co-authored-by: Luisa Santus <[email protected]>

modules/nf-core/kalign/align/environment.yml

@@ -4,4 +4,5 @@ channels:
   - bioconda
   - defaults
 dependencies:
-  - bioconda::kalign3=3.3.5
+  - bioconda::kalign3=3.4.0
+  - conda-forge::pigz=2.8

modules/nf-core/kalign/align/main.nf

@@ -4,43 +4,47 @@ process KALIGN_ALIGN {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'https://depot.galaxyproject.org/singularity/kalign3:3.3.5--hdbdd923_0':
-        'biocontainers/kalign3:3.3.5--hdbdd923_0' }"
+        'https://depot.galaxyproject.org/singularity/mulled-v2-5cd0277547c6b33133225c8ce14c0cf2a4396ea2:0a70b6d89a3e06fbdc4a735461e8b98ff32ee5de-0':
+        'biocontainers/mulled-v2-5cd0277547c6b33133225c8ce14c0cf2a4396ea2:0a70b6d89a3e06fbdc4a735461e8b98ff32ee5de-0' }"
 
     input:
     tuple val(meta), path(fasta)
+    val(compress)
 
     output:
-    tuple val(meta), path("*.aln"), emit: alignment
-    path "versions.yml"           , emit: versions
+    tuple val(meta), path("*.aln{.gz,}"), emit: alignment
+    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 write_output = compress ? ">(pigz -cp ${task.cpus} > ${prefix}.aln.gz)" : "${prefix}.aln"
     """
+    unpigz -cdf $fasta | \\
     kalign \\
         $args \\
-        -i $fasta \\
-        -o ${prefix}.aln
+        -o ${write_output}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
-        kalign : \$(echo \$(kalign -v) | sed 's/kalign //g' )
+        kalign: \$(echo \$(kalign -v) | sed 's/kalign //g' )
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
     END_VERSIONS
     """
 
     stub:
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
     """
-    touch ${prefix}.aln
+    touch ${prefix}.aln${compress ? '.gz' : ''}
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         kalign : \$(echo \$(kalign -v) | sed 's/kalign //g' )
+        pigz: \$(echo \$(pigz --version 2>&1) | sed 's/^.*pigz\\w*//' ))
     END_VERSIONS
     """
 }

modules/nf-core/kalign/align/meta.yml

@@ -20,8 +20,11 @@ input:
         e.g. `[ id:'test']`
   - fasta:
       type: file
-      description: Input sequences in FASTA format
-      pattern: "*.{fa,fasta}"
+      description: Input sequences in FASTA format. May be gzipped or uncompressed.
+      pattern: "*.{fa,fasta}{.gz,}"
+  - compress:
+      type: boolean
+      description: Flag representing whether the output MSA should be compressed. Set to true to enable/false to disable compression. Compression is done using pigz, and is multithreaded.
 output:
   - meta:
       type: map
@@ -30,8 +33,8 @@ output:
         e.g. `[ id:'test']`
   - alignment:
       type: file
-      description: Alignment file.
-      pattern: "*.{aln}"
+      description: Alignment file. May be gzipped or uncompressed, depending on if `compress` is set to `true` or `false`.
+      pattern: "*.{aln}{.gz,}"
   - versions:
       type: file
       description: File containing software versions

modules/nf-core/kalign/align/tests/main.nf.test

@@ -10,24 +10,45 @@ nextflow_process {
     tag "kalign"
     tag "kalign/align"
 
-    test("sarscov2 - fasta") {
+    test("sarscov2 - fasta - uncompressed") {
 
         when {
             process {
                 """
                 input[0] = [ [ id:'test' ], // meta map
                           file(params.test_data['sarscov2']['illumina']['contigs_fasta'], checkIfExists: true)
                         ]
+                input[1] = false
                 """
             }
         }
 
         then {
             assertAll(
                 { assert process.success },
-                { assert snapshot(process.out.alignment).match("alignment")},
-                { assert snapshot(process.out.versions).match("versions") }
+                { assert snapshot(process.out).match("SARS-CoV-2 scaffolds fasta - uncompressed")},
             )
         }
     }
-}
+
+    test("sarscov2 - fasta - compressed") {
+
+        when {
+            process {
+                """
+                input[0] = [ [ id:'test' ], // meta map
+                          file(params.test_data['sarscov2']['illumina']['contigs_fasta'], checkIfExists: true)
+                        ]
+                input[1] = true
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match("SARS-CoV-2 scaffolds fasta - compressed")},
+            )
+        }
+    }
+}