Fix: removes hardcoded flags and 'custom' file output for AdapterRemoval (#1357)

* fix: remove left-over unnecessary code

* Removes hardcoded flags and more explicit output

* Fix test md5

modules/adapterremoval/main.nf

@@ -11,9 +11,15 @@ process ADAPTERREMOVAL {
     tuple val(meta), path(reads)
 
     output:
-    tuple val(meta), path('*.fastq.gz'), emit: reads
+    tuple val(meta), path('*.truncated.gz')         , optional: true, emit: singles_truncated
-    tuple val(meta), path('*.log')     , emit: log
+    tuple val(meta), path('*.discarded.gz')         , optional: true, emit: discarded
-    path "versions.yml"                , emit: versions
+    tuple val(meta), path('*.pair1.truncated.gz')   , optional: true, emit: pair1_truncated
+    tuple val(meta), path('*.pair2.truncated.gz')   , optional: true, emit: pair2_truncated
+    tuple val(meta), path('*.collapsed.gz')         , optional: true, emit: collapsed
+    tuple val(meta), path('*.collapsed.truncated')  , optional: true, emit: collapsed_truncated
+    tuple val(meta), path('*paired.gz')             , optional: true, emit: paired_interleaved
+    tuple val(meta), path('*.log')                  , emit: log
+    path "versions.yml"                             , emit: versions
 
     when:
     task.ext.when == null || task.ext.when
@@ -28,30 +34,27 @@ process ADAPTERREMOVAL {
             --file1 $reads \\
             $args \\
             --basename $prefix \\
-            --threads $task.cpus \\
+            --threads ${task.cpus} \\
             --settings ${prefix}.log \\
-            --output1 ${prefix}.trimmed.fastq.gz \\
             --seed 42 \\
-            --gzip \\
+            --gzip
 
         cat <<-END_VERSIONS > versions.yml
         "${task.process}":
             adapterremoval: \$(AdapterRemoval --version 2>&1 | sed -e "s/AdapterRemoval ver. //g")
         END_VERSIONS
         """
-    } else if (!meta.single_end && !meta.collapse) {
+    } else if (!meta.single_end ) {
         """
         AdapterRemoval  \\
             --file1 ${reads[0]} \\
             --file2 ${reads[1]} \\
             $args \\
             --basename $prefix \\
-            --threads $task.cpus \\
+            --threads ${task.cpus} \\
             --settings ${prefix}.log \\
-            --output1 ${prefix}.pair1.trimmed.fastq.gz \\
-            --output2 ${prefix}.pair2.trimmed.fastq.gz \\
             --seed 42 \\
-            --gzip \\
+            --gzip
 
         cat <<-END_VERSIONS > versions.yml
         "${task.process}":
@@ -63,13 +66,12 @@ process ADAPTERREMOVAL {
         AdapterRemoval  \\
             --file1 ${reads[0]} \\
             --file2 ${reads[1]} \\
-            --collapse \\
             $args \\
             --basename $prefix \\
             --threads $task.cpus \\
             --settings ${prefix}.log \\
             --seed 42 \\
-            --gzip \\
+            --gzip
 
         cat *.collapsed.gz *.collapsed.truncated.gz > ${prefix}.merged.fastq.gz
         cat <<-END_VERSIONS > versions.yml

modules/adapterremoval/meta.yml

@@ -17,13 +17,13 @@ input:
       type: map
       description: |
         Groovy Map containing sample information
-        e.g. [ id:'test', single_end:false, collapse:false ]
+        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.
-      pattern: "*.{fq,fastq,fg.gz,fastq.gz}"
+      pattern: "*.{fq,fastq,fq.gz,fastq.gz}"
 
 output:
   - meta:
@@ -31,12 +31,45 @@ output:
       description: |
         Groovy Map containing sample information
         e.g. [ id:'test', single_end:false ]
-  - reads:
+  - singles_truncated:
       type: file
       description: |
-        List of input adapter trimmed FastQ files of size 1 or 2 for
+        Adapter trimmed FastQ files of either single-end reads, or singleton
-        single-end or collapsed data and paired-end data, respectively.
+        'orphaned' reads from merging of paired-end data (i.e., one of the pair
-      pattern: "*.{fastq.gz}"
+        was lost due to filtering thresholds).
+      pattern: "*.truncated.gz"
+  - discarded:
+      type: file
+      description: |
+        Adapter trimmed FastQ files of reads that did not pass filtering
+        thresholds.
+      pattern: "*.discarded.gz"
+  - pair1_truncated:
+      type: file
+      description: |
+        Adapter trimmed R1 FastQ files of paired-end reads that did not merge
+        with their respective R2 pair due to long templates. The respective pair
+        is stored in 'pair2_truncated'.
+      pattern: "*.pair1.truncated.gz"
+  - pair2_truncated:
+      type: file
+      description: |
+        Adapter trimmed R2 FastQ files of paired-end reads that did not merge
+        with their respective R1 pair due to long templates. The respective pair
+        is stored in 'pair1_truncated'.
+      pattern: "*.pair2.truncated.gz"
+  - collapsed:
+      type: file
+      description: |
+        Collapsed FastQ of paired-end reads that successfully merged with their
+        respective R1 pair but were not trimmed.
+      pattern: "*.collapsed.gz"
+  - collapsed_truncated:
+      type: file
+      description: |
+        Collapsed FastQ of paired-end reads that successfully merged with their
+        respective R1 pair and were trimmed of adapter due to sufficient overlap.
+      pattern: "*.collapsed.truncated.gz"
   - log:
       type: file
       description: AdapterRemoval log file
@@ -48,3 +81,4 @@ output:
 
 authors:
   - "@maxibor"
+  - "@jfy133"

tests/modules/adapterremoval/test.yml

@@ -1,31 +1,44 @@
 - name: adapterremoval test_adapterremoval_single_end
-  command: nextflow run ./tests/modules/adapterremoval -entry test_adapterremoval_single_end -c ./tests/config/nextflow.config -c ./tests/modules/adapterremoval/nextflow.config
+  command: nextflow run tests/modules/adapterremoval -entry test_adapterremoval_single_end -c tests/config/nextflow.config
   tags:
     - adapterremoval
   files:
+    - path: output/adapterremoval/test.discarded.gz
     - path: output/adapterremoval/test.log
       md5sum: 2fd3d5d703b63ba33a83021fccf25f77
-    - path: output/adapterremoval/test.trimmed.fastq.gz
+    - path: output/adapterremoval/test.truncated.gz
       md5sum: 62139afee94defad5b83bdd0b8475a1f
+    - path: output/adapterremoval/versions.yml
+      md5sum: ac5b46719719b7ee62739530b80869fc
 
 - name: adapterremoval test_adapterremoval_paired_end
-  command: nextflow run ./tests/modules/adapterremoval -entry test_adapterremoval_paired_end -c ./tests/config/nextflow.config -c ./tests/modules/adapterremoval/nextflow.config
+  command: nextflow run tests/modules/adapterremoval -entry test_adapterremoval_paired_end -c tests/config/nextflow.config
   tags:
     - adapterremoval
   files:
+    - path: output/adapterremoval/test.discarded.gz
     - path: output/adapterremoval/test.log
       md5sum: b8a451d3981b327f3fdb44f40ba2d6d1
-    - path: output/adapterremoval/test.pair1.trimmed.fastq.gz
+    - path: output/adapterremoval/test.pair1.truncated.gz
       md5sum: 294a6277f0139bd597e57c6fa31f39c7
-    - path: output/adapterremoval/test.pair2.trimmed.fastq.gz
+    - path: output/adapterremoval/test.pair2.truncated.gz
       md5sum: de7b38e2c881bced8671acb1ab452d78
+    - path: output/adapterremoval/test.singleton.truncated.gz
+    - path: output/adapterremoval/versions.yml
+      md5sum: fa621c887897da5a379c719399c17db7
 
 - name: adapterremoval test_adapterremoval_paired_end_collapse
-  command: nextflow run ./tests/modules/adapterremoval -entry test_adapterremoval_paired_end_collapse -c ./tests/config/nextflow.config -c ./tests/modules/adapterremoval/nextflow.config
+  command: nextflow run tests/modules/adapterremoval -entry test_adapterremoval_paired_end_collapse -c tests/config/nextflow.config
   tags:
     - adapterremoval
   files:
+    - path: output/adapterremoval/test.discarded.gz
     - path: output/adapterremoval/test.log
-      md5sum: 7f0b2328152226e46101a535cce718b3
+      md5sum: b8a451d3981b327f3fdb44f40ba2d6d1
-    - path: output/adapterremoval/test.merged.fastq.gz
+    - path: output/adapterremoval/test.pair1.truncated.gz
-      md5sum: 07a8f725bfd3ecbeabdc41b32d898dee
+      md5sum: 294a6277f0139bd597e57c6fa31f39c7
+    - path: output/adapterremoval/test.pair2.truncated.gz
+      md5sum: de7b38e2c881bced8671acb1ab452d78
+    - path: output/adapterremoval/test.singleton.truncated.gz
+    - path: output/adapterremoval/versions.yml
+      md5sum: fd428f92a8446e0b34c5ae1c447215b8