Adding sentieon-dedup

conf/modules/modules.config

@@ -40,7 +40,7 @@ process {
         ]
     }
 
-    withName: 'NFCORE_SAREK:SAREK:(BAM_MARKDUPLICATES|BAM_MARKDUPLICATES_SPARK):CRAM_QC_MOSDEPTH_SAMTOOLS:SAMTOOLS_STATS' {
+    withName: 'NFCORE_SAREK:SAREK:(BAM_MARKDUPLICATES|BAM_MARKDUPLICATES_SPARK|BAM_SENTIEON_DEDUP):CRAM_QC_MOSDEPTH_SAMTOOLS:SAMTOOLS_STATS' {
         ext.when         = { !(params.skip_tools && params.skip_tools.split(',').contains('samtools')) }
         ext.prefix       = { "${meta.id}.md.cram" }
         publishDir       = [

conf/modules/sentieon_dedup.config

@@ -0,0 +1,46 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    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.
+----------------------------------------------------------------------------------------
+*/
+
+// SENTIEON_DEDUP
+
+process {
+
+    withName: 'SENTIEON_DEDUP' {
+        ext.prefix       = { "${meta.id}.md" }
+        ext.when         = { params.tools && params.tools.split(',').contains('sentieon_dedup') }
+        publishDir       = [
+            [
+                enabled: !params.save_output_as_bam,
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/preprocessing/sentieon_dedup/${meta.id}/" },
+                pattern: "*{cram,crai}"
+            ],
+            [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/reports/" },
+                pattern: "*metrics",
+                saveAs: { !(params.skip_tools && params.skip_tools.split(',').contains('sentieon_dedup_report')) ? "sentieon_dedup/${meta.id}/${it}" : null}
+            ]
+        ]
+    }
+
+    withName: 'INDEX_DEDUPED' {
+        publishDir       = [
+            enabled: !params.save_output_as_bam,
+            mode: params.publish_dir_mode,
+            path: { "${params.outdir}/preprocessing/sentieon_dedup/${meta.id}/" },
+            pattern: "*{cram,crai}"
+        ]
+    }
+
+}

conf/test/test_sentieon_dedup_bam.config

@@ -0,0 +1,16 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Nextflow config file for running minimal tests
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Defines input files and everything required to run a fast and simple pipeline test.
+
+    Use as follows:
+        nextflow run nf-core/sarek -profile test,<extra_test_profile>,<docker/singularity> --outdir <OUTDIR>
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+*/
+
+params {
+    input = "${projectDir}/tests/csv/3.0/mapped_single_bam.csv"
+    step  = 'markduplicates'
+    tools = 'sentieon_dedup'
+}

conf/test/test_sentieon_dedup_cram.config

@@ -0,0 +1,16 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Nextflow config file for running minimal tests
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    Defines input files and everything required to run a fast and simple pipeline test.
+
+    Use as follows:
+        nextflow run nf-core/sarek -profile test,<extra_test_profile>,<docker/singularity> --outdir <OUTDIR>
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+*/
+
+params {
+    input = "${projectDir}/tests/csv/3.0/mapped_single_cram.csv"
+    step  = 'markduplicates'
+    tools = 'sentieon_dedup'
+}

modules.json

@@ -378,6 +378,11 @@
                         "git_sha": "8c159ebb28cfa08890042a133bfab78de932eff7",
                         "installed_by": ["modules"]
                     },
+                    "sentieon/dedup": {
+                        "branch": "master",
+                        "git_sha": "8c04ecc0c5d4ad102755b919b2525481f0380090",
+                        "installed_by": ["modules"]
+                    },
                     "snpeff/download": {
                         "branch": "master",
                         "git_sha": "aeefdd170be4583012816212f1345fbebb61ccd6",

nextflow.config

@@ -218,6 +218,8 @@ profiles {
     annotation                 { includeConfig 'conf/test/test_annotation.config' }
     markduplicates_bam         { includeConfig 'conf/test/test_markduplicates_bam.config' }
     markduplicates_cram        { includeConfig 'conf/test/test_markduplicates_cram.config' }
+    sentieon_dedup_bam         { includeConfig 'conf/test/test_sentieon_dedup_bam.config' }
+    sentieon_dedup_cram        { includeConfig 'conf/test/test_sentieon_dedup_cram.config' }
     no_intervals               { includeConfig 'conf/test/test_no_intervals.config' }
     pair                       { includeConfig 'conf/test/test_pair.config' }
     prepare_recalibration_bam  { includeConfig 'conf/test/test_prepare_recalibration_bam.config' }
@@ -304,6 +306,7 @@ includeConfig 'conf/modules/prepare_intervals.config'
 includeConfig 'conf/modules/aligner.config'
 includeConfig 'conf/modules/alignment_to_fastq.config'
 includeConfig 'conf/modules/markduplicates.config'
+includeConfig 'conf/modules/sentieon_dedup.config'
 includeConfig 'conf/modules/prepare_recalibration.config'
 includeConfig 'conf/modules/recalibrate.config'
 includeConfig 'conf/modules/trimming.config'

nextflow_schema.json

@@ -89,7 +89,7 @@
                     "fa_icon": "fas fa-toolbox",
                     "description": "Tools to use for variant calling and/or for annotation.",
                     "help_text": "Multiple tools separated with commas.\n\n**Variant Calling:**\n\nGermline variant calling can currently be performed with the following variant callers:\n- SNPs/Indels: DeepVariant, FreeBayes, HaplotypeCaller, mpileup, Strelka\n- Structural Variants: Manta, TIDDIT\n- Copy-number: CNVKit\n\nTumor-only somatic variant calling can currently be performed with the following variant callers:\n- SNPs/Indels: FreeBayes, mpileup, Mutect2, Strelka\n- Structural Variants: Manta, TIDDIT\n- Copy-number: CNVKit, ControlFREEC\n\nSomatic variant calling can currently only be performed with the following variant callers:\n- SNPs/Indels: FreeBayes, Mutect2, Strelka2\n- Structural variants: Manta, TIDDIT\n- Copy-Number: ASCAT, CNVKit, Control-FREEC\n- Microsatellite Instability: MSIsensorpro\n\n> **NB** Mutect2 for somatic variant calling cannot be combined with `--no_intervals`\n\n**Annotation:**\n \n- snpEff, VEP, merge (both consecutively).\n\n> **NB** As Sarek will use bgzip and tabix to compress and index VCF files annotated, it expects VCF files to be sorted when starting from `--step annotate`.",
-                    "pattern": "^((ascat|cnvkit|controlfreec|deepvariant|freebayes|haplotypecaller|manta|merge|mpileup|msisensorpro|mutect2|snpeff|strelka|tiddit|vep)?,?)*(?<!,)$"
+                    "pattern": "^((ascat|cnvkit|controlfreec|deepvariant|freebayes|haplotypecaller|manta|merge|mpileup|msisensorpro|mutect2|sentieon_dedup|snpeff|strelka|tiddit|vep)?,?)*(?<!,)$"
                 },
                 "skip_tools": {
                     "type": "string",

subworkflows/local/bam_sentieon_dedup/main.nf

@@ -0,0 +1,53 @@
+//
+// SENTIEON DEDUP
+//
+// For all modules here:
+// A when clause condition is defined in the conf/modules.config to determine if the module should be run
+
+include { CRAM_QC_MOSDEPTH_SAMTOOLS              } from '../cram_qc_mosdepth_samtools/main'
+include { GATK4_MARKDUPLICATES                   } from '../../../modules/nf-core/gatk4/markduplicates/main'
+include { SENTIEON_DEDUP                         } from '../../../modules/nf-core/sentieon/dedup/main'
+include { SAMTOOLS_INDEX as INDEX_INPUT          } from '../../../modules/nf-core/samtools/index/main'
+include { SAMTOOLS_INDEX as INDEX_DEDUPED        } from '../../../modules/nf-core/samtools/index/main'
+include { SAMTOOLS_INDEX as INDEX_MARKDUPLICATES } from '../../../modules/nf-core/samtools/index/main'
+
+workflow BAM_SENTIEON_DEDUP {
+    take:
+    bam                    // channel: [mandatory] [ meta, bam ]  // Although the channel is named "bam", it may contain cram-files.
+    fasta                  // channel: [mandatory] [ fasta ]
+    fasta_fai              // channel: [mandatory] [ fasta_fai ]
+    intervals_bed_combined // channel: [optional]  [ intervals_bed ]
+
+    main:
+    versions = Channel.empty()
+    reports  = Channel.empty()
+
+    INDEX_INPUT(bam)
+    bam_bai = bam.join(INDEX_INPUT.out.bai.concat(INDEX_INPUT.out.crai), failOnMismatch:true, failOnDuplicate:true)
+    // The concat operation is part of the above command since if the "bam" channel contains cram-files, then the index files will be in the channel INDEX_INPUT.out.crai and not in INDEX_INPUT.out.bai
+    SENTIEON_DEDUP(bam_bai, fasta, fasta_fai)
+
+    INDEX_DEDUPED(SENTIEON_DEDUP.out.cram)
+
+    // Join with the crai file
+    cram = SENTIEON_DEDUP.out.cram.join(INDEX_DEDUPED.out.crai, failOnDuplicate: true, failOnMismatch: true)
+
+    // QC on CRAM
+    CRAM_QC_MOSDEPTH_SAMTOOLS(cram, fasta, intervals_bed_combined)
+
+    // Gather all reports generated
+    reports = reports.mix(SENTIEON_DEDUP.out.metrics)
+    reports = reports.mix(SENTIEON_DEDUP.out.score)
+    reports = reports.mix(CRAM_QC_MOSDEPTH_SAMTOOLS.out.reports)
+
+    // Gather versions of all tools used
+    versions = versions.mix(SENTIEON_DEDUP.out.versions)
+    versions = versions.mix(INDEX_DEDUPED.out.versions)
+    versions = versions.mix(CRAM_QC_MOSDEPTH_SAMTOOLS.out.versions)
+
+    emit:
+    cram
+    reports
+
+    versions    // channel: [ versions.yml ]
+}