switch to using --write-index instead of samtools index

CHANGELOG.md

@@ -14,6 +14,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Changed
 
+- [#1338](https://github.com/nf-core/sarek/pull/1338) - Optimisation: switch `samtools index` to `--write-index` ([#1121](https://github.com/nf-core/sarek/issues/1121) )
+
 ### Fixed
 
 - [#1334](https://github.com/nf-core/sarek/pull/1334) - Remove extra v, when reporting tower runs on slack

conf/modules/aligner.config

@@ -80,6 +80,7 @@ process {
     }
 
     withName: 'MERGE_BAM' {
+        ext.args   = { "--write-index" }
         ext.prefix = { "${meta.id}.sorted" }
     }
 }

conf/modules/haplotypecaller.config

@@ -63,5 +63,9 @@ process {
                 pattern: "*{bam,bai}"
             ]
         }
+
+        withName: 'NFCORE_SAREK:SAREK:BAM_VARIANT_CALLING_GERMLINE_ALL:BAM_VARIANT_CALLING_HAPLOTYPECALLER:BAM_MERGE_INDEX_SAMTOOLS:MERGE_BAM' {
+            ext.args = { "--write-index" }
+        }
     }
 }

conf/modules/ngscheckmate.config

@@ -1,3 +1,16 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Config file for defining DSL2 per module options and publishing paths
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Available keys to override module options:
+        ext.args   = Additional arguments appended to command in module.
+        ext.args2  = Second set of arguments appended to command in module (multi-tool modules).
+        ext.args3  = Third set of arguments appended to command in module (multi-tool modules).
+        ext.prefix = File name prefix for output files.
+        ext.when   = When to run the module.
+----------------------------------------------------------------------------------------
+*/
+
 process {
 
     withName: '.*BAM_NGSCHECKMATE:BCFTOOLS_MPILEUP' {

conf/modules/recalibrate.config

@@ -27,19 +27,20 @@ process {
 
     if ((params.step == 'mapping' || params.step == 'markduplicates'|| params.step == 'prepare_recalibration'|| params.step == 'recalibrate') && (!(params.skip_tools && params.skip_tools.split(',').contains('baserecalibrator')))) {
         withName: 'NFCORE_SAREK:SAREK:(BAM_APPLYBQSR|BAM_APPLYBQSR_SPARK):CRAM_MERGE_INDEX_SAMTOOLS:MERGE_CRAM' {
+            ext.args         = { "--write-index"    }
             ext.prefix       = { "${meta.id}.recal" }
             ext.when         = { meta.num_intervals > 1 }
             publishDir       = [
-                    mode: params.publish_dir_mode,
+                mode: params.publish_dir_mode,
                 path: { "${params.outdir}/preprocessing/recalibrated/${meta.id}/" },
-                pattern: "*cram",
+                pattern: "*cram,*cram.crai",
                 saveAs: { !params.save_output_as_bam ? it : null }
         ]
         }
 
         withName: 'NFCORE_SAREK:SAREK:(BAM_APPLYBQSR|BAM_APPLYBQSR_SPARK):CRAM_MERGE_INDEX_SAMTOOLS:INDEX_CRAM' {
             publishDir       = [
-                    mode: params.publish_dir_mode,
+                mode: params.publish_dir_mode,
                 path: { "${params.outdir}/preprocessing/recalibrated/${meta.id}/" },
                 pattern: "*{recal.cram,recal.cram.crai}",
                 saveAs: { !params.save_output_as_bam ? it : null }

subworkflows/local/bam_merge_index_samtools/main.nf

@@ -25,14 +25,15 @@ workflow BAM_MERGE_INDEX_SAMTOOLS {
     // Only when using intervals
     MERGE_BAM(bam_to_merge.multiple, [ [ id:'null' ], []], [ [ id:'null' ], []])
 
-    // Mix intervals and no_intervals channels together
-    bam_all = MERGE_BAM.out.bam.mix(bam_to_merge.single)
+    // Join with the bai file
+    bam_bai_merged = MERGE_BAM.out.bam.join(MERGE_BAM.out.bai)
 
-    // Index bam
-    INDEX_MERGE_BAM(bam_all)
+    // Index single bams, merged ones are indexed on the fly
+    INDEX_MERGE_BAM(bam_to_merge.single)
+    bam_bai_single = bam_to_merge.single.join(INDEX_MERGE_BAM.out.bai, failOnDuplicate: true, failOnMismatch: true)
 
-    // Join with the bai file
-    bam_bai = bam_all.join(INDEX_MERGE_BAM.out.bai, failOnDuplicate: true, failOnMismatch: true)
+    // Mix intervals and no_intervals channels together
+    bam_bai = bam_bai_merged.mix(bam_bai_single)
 
     // Gather versions of all tools used
     versions = versions.mix(INDEX_MERGE_BAM.out.versions)

subworkflows/local/cram_merge_index_samtools/main.nf

@@ -27,14 +27,15 @@ workflow CRAM_MERGE_INDEX_SAMTOOLS {
     // Only when using intervals
     MERGE_CRAM(cram_to_merge.multiple, fasta.map{ it -> [ [ id:'fasta' ], it ] }, fasta_fai.map{ it -> [ [ id:'fasta_fai' ], it ] })
 
-    // Mix intervals and no_intervals channels together
-    cram_all = MERGE_CRAM.out.cram.mix(cram_to_merge.single)
+    // Join with the crai file
+    cram_crai_merged = MERGE_CRAM.out.cram.join(MERGE_CRAM.out.crai)
 
-    // Index cram
-    INDEX_CRAM(cram_all)
+    // Index cram, multiple ones are indexed on the fly
+    INDEX_CRAM(cram_to_merge.single)
+    cram_crai_single = cram_to_merge.single.join(INDEX_CRAM.out.crai)
 
-    // Join with the crai file
-    cram_crai = cram_all.join(INDEX_CRAM.out.crai, failOnDuplicate: true, failOnMismatch: true)
+    // Mix intervals and no_intervals channels together
+    cram_crai = cram_crai_merged.mix(cram_crai_single)
 
     // Gather versions of all tools used
     versions = versions.mix(INDEX_CRAM.out.versions.first())