Merge pull request #558 from nf-core/fix-cat-se-data-dev

Fix malformed input to MEGAHIT with non-run-merged SEdata
nf-core · Jan 26, 2024 · eb97cbd · eb97cbd
2 parents 13212ab + ff0c4aa
commit eb97cbd
Show file tree

Hide file tree

Showing 4 changed files with 21 additions and 11 deletions.
diff --git a/.devcontainer/devcontainer.json b/.devcontainer/devcontainer.json
@@ -18,11 +18,11 @@
                 "python.linting.flake8Path": "/opt/conda/bin/flake8",
                 "python.linting.pycodestylePath": "/opt/conda/bin/pycodestyle",
                 "python.linting.pydocstylePath": "/opt/conda/bin/pydocstyle",
-                "python.linting.pylintPath": "/opt/conda/bin/pylint"
+                "python.linting.pylintPath": "/opt/conda/bin/pylint",
             },
 
             // Add the IDs of extensions you want installed when the container is created.
-            "extensions": ["ms-python.python", "ms-python.vscode-pylance", "nf-core.nf-core-extensionpack"]
-        }
-    }
+            "extensions": ["ms-python.python", "ms-python.vscode-pylance", "nf-core.nf-core-extensionpack"],
+        },
+    },
 }
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -11,6 +11,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### `Fixed`
 
+- [#558](https://github.com/nf-core/mag/pull/558) - Fix bug in run merging when dealing with single end data (reported by @roberta-davidson, fix by @jfy133)
+
 ### `Dependencies`
 
 ### `Deprecated`

diff --git a/subworkflows/local/ancient_dna.nf b/subworkflows/local/ancient_dna.nf
@@ -29,9 +29,9 @@ workflow ANCIENT_DNA_ASSEMBLY_VALIDATION {
                                         fasta: [ contigs ]
                                         fai: [ fai ]
                                 }
-            FREEBAYES ( freebayes_input.reads.dump(tag: 'reads'),
-                        freebayes_input.fasta.dump(tag: 'fasta'),
-                        freebayes_input.fai.dump(tag: 'fai'),
+            FREEBAYES ( freebayes_input.reads,
+                        freebayes_input.fasta,
+                        freebayes_input.fai,
                         [],
                         [],
                         [] )

diff --git a/workflows/mag.nf b/workflows/mag.nf
@@ -350,15 +350,23 @@ workflow MAG {
             .groupTuple()
             .branch {
                 meta, reads ->
-                    cat:       ( meta.single_end && reads.size() == 1 ) || ( !meta.single_end && reads.size() >= 2 )
+                    cat:      reads.size() >= 2 // SE: [[meta], [S1_R1, S2_R1]]; PE: [[meta], [[S1_R1, S1_R2], [S2_R1, S2_R2]]]
                     skip_cat: true // Can skip merging if only single lanes
             }
 
         CAT_FASTQ ( ch_short_reads_forcat.cat.map { meta, reads -> [ meta, reads.flatten() ]} )
 
-            ch_short_reads = Channel.empty()
-            ch_short_reads = CAT_FASTQ.out.reads.mix( ch_short_reads_forcat.skip_cat ).map { meta, reads -> [ meta, reads.flatten() ]}
-            ch_versions    = ch_versions.mix(CAT_FASTQ.out.versions.first())
+        // Ensure we don't have nests of nests so that structure is in form expected for assembly
+        ch_short_reads_catskipped = ch_short_reads_forcat.skip_cat
+                                        .map { meta, reads ->
+                                            def new_reads = meta.single_end ? reads[0] : reads.flatten()
+                                        [ meta, new_reads ]
+                                    }
+
+        // Combine single run and multi-run-merged data
+        ch_short_reads = Channel.empty()
+        ch_short_reads = CAT_FASTQ.out.reads.mix(ch_short_reads_catskipped)
+        ch_versions    = ch_versions.mix(CAT_FASTQ.out.versions.first())
 
         if ( params.bbnorm ) {
             if ( params.coassemble_group ) {