Merge pull request #162 from genomic-medicine-sweden/short_reads_samtools

Add samtools stats for short-reads

README.md

@@ -34,9 +34,10 @@ On release, automated continuous integration tests run the pipeline on a full-si
 2. Performs optional read pre-processing
    - Adapter clipping and merging (short-read: [fastp](https://github.com/OpenGene/fastp), [AdapterRemoval2](https://github.com/MikkelSchubert/adapterremoval); long-read: [porechop](https://github.com/rrwick/Porechop))
    - Low complexity and quality filtering (short-read: [bbduk](https://jgi.doe.gov/data-and-tools/software-tools/bbtools/), [PRINSEQ++](https://github.com/Adrian-Cantu/PRINSEQ-plus-plus); long-read: [Filtlong](https://github.com/rrwick/Filtlong))
-   - Host read removal (short-read: [BowTie2](http://bowtie-bio.sourceforge.net/bowtie2/); long-read: [Minimap2](https://github.com/lh3/minimap2))
+   - Host-read removal (short-read: [BowTie2](http://bowtie-bio.sourceforge.net/bowtie2/); long-read: [Minimap2](https://github.com/lh3/minimap2))
    - Run merging
-3. Performs taxonomic profiling using one or more of:
+3. Supports statistics for host-read removal ([Samtools](http://www.htslib.org/))
+4. Performs taxonomic profiling using one or more of:
    - [Kraken2](https://ccb.jhu.edu/software/kraken2/)
    - [MetaPhlAn3](https://huttenhower.sph.harvard.edu/metaphlan/)
    - [MALT](https://uni-tuebingen.de/fakultaeten/mathematisch-naturwissenschaftliche-fakultaet/fachbereiche/informatik/lehrstuehle/algorithms-in-bioinformatics/software/malt/)
@@ -45,11 +46,11 @@ On release, automated continuous integration tests run the pipeline on a full-si
    - [Kaiju](https://kaiju.binf.ku.dk/)
    - [mOTUs](https://motu-tool.org/)
    - [KrakenUniq](https://github.com/fbreitwieser/krakenuniq)
-4. Perform optional post-processing with:
+5. Perform optional post-processing with:
    - [bracken](https://ccb.jhu.edu/software/bracken/)
-5. Standardises output tables
+6. Standardises output tables
-6. Present QC for raw reads ([`MultiQC`](http://multiqc.info/))
+7. Present QC for raw reads ([`MultiQC`](http://multiqc.info/))
-7. Plotting Kraken2, Centrifuge, Kaiju and MALT results ([`Krona`](https://hpc.nih.gov/apps/kronatools.html))
+8. Plotting Kraken2, Centrifuge, Kaiju and MALT results ([`Krona`](https://hpc.nih.gov/apps/kronatools.html))
 
 ## Quick Start
 

conf/modules.config

@@ -236,7 +236,6 @@ process {
         publishDir = [
             path: { "${params.outdir}/samtools/stats" },
             mode: params.publish_dir_mode,
-            enabled: params.save_hostremoval_unmapped,
             pattern: '*stats'
         ]
     }

subworkflows/local/shortread_hostremoval.nf

@@ -4,6 +4,9 @@
 
 include { BOWTIE2_BUILD             } from '../../modules/nf-core/bowtie2/build/main'
 include { BOWTIE2_ALIGN             } from '../../modules/nf-core/bowtie2/align/main'
+include { SAMTOOLS_INDEX            } from '../../modules/nf-core/samtools/index/main'
+include { SAMTOOLS_STATS            } from '../../modules/nf-core/samtools/stats/main'
+include { SAMTOOLS_VIEW             } from '../../modules/nf-core/samtools/view/main'
 
 workflow SHORTREAD_HOSTREMOVAL {
     take:
@@ -26,7 +29,27 @@ workflow SHORTREAD_HOSTREMOVAL {
     ch_versions      = ch_versions.mix( BOWTIE2_ALIGN.out.versions.first() )
     ch_multiqc_files = ch_multiqc_files.mix( BOWTIE2_ALIGN.out.log )
 
+    ch_bowtie2_mapped = BOWTIE2_ALIGN.out.bam
+        .map {
+            meta, reads ->
+                [ meta, reads, [] ]
+        }
+
+    SAMTOOLS_VIEW ( ch_bowtie2_mapped, [], [] )
+    ch_versions      = ch_versions.mix( SAMTOOLS_VIEW.out.versions.first() )
+
+    SAMTOOLS_INDEX ( SAMTOOLS_VIEW.out.bam )
+    ch_versions      = ch_versions.mix( SAMTOOLS_INDEX.out.versions.first() )
+
+    bam_bai = SAMTOOLS_VIEW.out.bam
+        .join(SAMTOOLS_INDEX.out.bai, remainder: true)
+
+    SAMTOOLS_STATS ( bam_bai, reference )
+    ch_versions = ch_versions.mix(SAMTOOLS_STATS.out.versions.first())
+    ch_multiqc_files = ch_multiqc_files.mix( SAMTOOLS_STATS.out.stats )
+
     emit:
+    stats    = SAMTOOLS_STATS.out.stats
     reads    = BOWTIE2_ALIGN.out.fastq   // channel: [ val(meta), [ reads ] ]
     versions = ch_versions               // channel: [ versions.yml ]
     mqc      = ch_multiqc_files