Bug fix for pair end sharding and Formating changes for prettify

nf-core · Aug 25, 2023 · d7a711d · d7a711d
1 parent 21a246d
commit d7a711d
Show file tree

Hide file tree

Showing 4 changed files with 22 additions and 25 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,7 +7,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### `Added`
 
-- [#1006](https://github.com/nf-core/eager/issues/1006) Added feature to shard fastqs before mapping, allowing more flexibility in parallelisation of mapping. 
+- [#1006](https://github.com/nf-core/eager/issues/1006) Added feature to shard fastqs before mapping, allowing more flexibility in parallelisation of mapping.
 
 ### `Fixed`
 

diff --git a/CITATIONS.md b/CITATIONS.md
@@ -32,7 +32,8 @@
 
   > Broad Institute (2019). Picard Toolkit. GitHub Repository: https://broadinstitute.github.io/picard/
 
-- [SeqKit](https://bioinf.shenwei.me/seqkit/)  
+- [SeqKit](https://bioinf.shenwei.me/seqkit/)
+
   > Shen, W., Le, S., Li, Y., & Hu, F. (2016). SeqKit: A Cross-Platform and Ultrafast Toolkit for FASTA/Q File Manipulation. PLOS ONE, 11(10), e0163962. doi:[10.1371/journal.pone.0163962](https://doi.org/10.1371/journal.pone.0163962)
 
 - [bwa](https://doi.org/10.1093/bioinformatics/btp324)

diff --git a/conf/test.config b/conf/test.config
@@ -30,7 +30,7 @@ params {
     // Sharding FASTQ
     shard_fastq                           = true
     shard_size                             = 5000
-    
+
     // BAM filtering
     run_bamfiltering                      = true
     bamfiltering_minreadlength            = 30

diff --git a/subworkflows/local/map.nf b/subworkflows/local/map.nf
@@ -6,7 +6,7 @@ include { SEQKIT_SPLIT2
 include { FASTQ_ALIGN_BWAALN                                                                                                  } from '../../subworkflows/nf-core/fastq_align_bwaaln/main'
 include { BWA_MEM                                                                                                             } from '../../modules/nf-core/bwa/mem/main'
 include { BOWTIE2_ALIGN                                                                                                       } from '../../modules/nf-core/bowtie2/align/main'
-include { SAMTOOLS_MERGE as SAMTOOLS_MERGE_LANES ; SAMTOOLS_MERGE as SAMTOOLS_MERGE_SHARDS                                    } from '../../modules/nf-core/samtools/merge/main'
+include { SAMTOOLS_MERGE as SAMTOOLS_MERGE_LANES                                                                              } from '../../modules/nf-core/samtools/merge/main'
 include { SAMTOOLS_SORT  as SAMTOOLS_SORT_MERGED_LANES                                                                        } from '../../modules/nf-core/samtools/sort/main'
 include { SAMTOOLS_INDEX as SAMTOOLS_INDEX_MEM; SAMTOOLS_INDEX as SAMTOOLS_INDEX_BT2; SAMTOOLS_INDEX as SAMTOOLS_INDEX_MERGED_LANES } from '../../modules/nf-core/samtools/index/main'
 include { SAMTOOLS_FLAGSTAT as SAMTOOLS_FLAGSTAT_MAPPED                                                                       } from '../../modules/nf-core/samtools/flagstat/main'
@@ -27,32 +27,28 @@ workflow MAP {
         SEQKIT_SPLIT2( ch_input_for_sharding )
         ch_versions        = ch_versions.mix ( SEQKIT_SPLIT2.out.versions.first() )
 
-        ch_input_for_mapping = SEQKIT_SPLIT2.out.reads
+        reads = SEQKIT_SPLIT2.out.reads
             .transpose()
-            .combine(index)
-            .multiMap {
-                meta, reads, meta2, index ->
+            .map {
+                meta, reads ->
                     new_meta = meta.clone()
-                    new_meta.shard_number = reads.getName().replaceAll(/.*(part_\d+).fastq.gz/, '$1')
-                    new_meta.reference = meta2.id
-                    reads: [ new_meta, reads ] 
-                    index: [ meta2, index ]
-            } 
-
-    } else {
-
-        ch_input_for_mapping = reads
-                            .combine(index)
-                            .multiMap {
-                                meta, reads, meta2, index ->
-                                    new_meta = meta.clone()
-                                    new_meta.reference = meta2.id
-                                    reads: [ new_meta, reads ]
-                                    index: [ meta2, index]
-                            }
+                    new_meta.shard_number = reads.getName().replaceAll(/.*(part_\d+).(?:fastq|fq).gz/, '$1')
+                    [ new_meta, reads ] 
+            }
+            .groupTuple()
 
     }
 
+    ch_input_for_mapping = reads
+        .combine(index)
+        .multiMap {
+            meta, reads, meta2, index ->
+                new_meta = meta.clone()
+                new_meta.reference = meta2.id  
+                reads: [ new_meta, reads ] 
+                index: [ meta2, index ]
+        }
+
     if ( params.mapping_tool == 'bwaaln' ) {
         FASTQ_ALIGN_BWAALN ( ch_input_for_mapping.reads, ch_input_for_mapping.index )