WIP: Add sentieon tnseq

conf/modules/prepare_genome.config

@@ -107,7 +107,7 @@ process {
     }
 
     withName: 'TABIX_DBSNP' {
-        ext.when         = { !params.dbsnp_tbi && params.dbsnp && ((params.step == "mapping" || params.step == "markduplicates" || params.step == "prepare_recalibration") || params.tools && (params.tools.split(',').contains('controlfreec') || params.tools.split(',').contains('haplotypecaller') || params.tools.split(',').contains('sentieon_haplotyper') || params.tools.split(',').contains('sentieon_dnascope') || params.tools.split(',').contains('mutect2'))) }
+        ext.when         = { !params.dbsnp_tbi && params.dbsnp && ((params.step == "mapping" || params.step == "markduplicates" || params.step == "prepare_recalibration") || params.tools && (params.tools.split(',').contains('controlfreec') || params.tools.split(',').contains('haplotypecaller') || params.tools.split(',').contains('sentieon_haplotyper') || params.tools.split(',').contains('sentieon_dnascope') || params.tools.split(',').contains('mutect2') || params.tools.split(',').contains('sentieon_tnhaplotyper2'))) }
         publishDir       = [
             mode: params.publish_dir_mode,
             path: { "${params.outdir}/reference/dbsnp" },
@@ -117,7 +117,7 @@ process {
     }
 
     withName: 'TABIX_GERMLINE_RESOURCE' {
-        ext.when         = { !params.germline_resource_tbi && params.germline_resource && params.tools && params.tools.split(',').contains('mutect2') }
+        ext.when         = { !params.germline_resource_tbi && params.germline_resource && params.tools && (params.tools.split(',').contains('mutect2') || params.tools.split(',').contains('sentieon_tnhaplotyper2')) }
         publishDir       = [
             mode: params.publish_dir_mode,
             path: { "${params.outdir}/reference/germline_resource" },
@@ -147,7 +147,7 @@ process {
     }
 
     withName: 'TABIX_PON' {
-        ext.when         = { !params.pon_tbi && params.pon && params.tools && params.tools.split(',').contains('mutect2') }
+        ext.when         = { !params.pon_tbi && params.pon && params.tools && (params.tools.split(',').contains('mutect2') || params.tools.split(',').contains('sentieon_tnhaplotyper2')) }
         publishDir       = [
             mode: params.publish_dir_mode,
             path: { "${params.outdir}/reference/pon" },

conf/modules/sentieon_tnhaplotyper2.config

@@ -0,0 +1,116 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    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.
+----------------------------------------------------------------------------------------
+*/
+
+// TNHAPLOTYPER2
+
+process {
+
+    if (params.tools && (params.tools.split(',').contains('sentieon_tnhaplotyper2'))) {
+        withName: 'SENTIEON_TNHAPLOTYPER2' {
+            ext.prefix       = { meta.num_intervals <= 1 ? "${meta.id}.tnhaplotyper2" : "${meta.id}.tnhaplotyper2.${intervals.simpleName}" }
+            ext.when         = { params.tools && (params.tools.split(',').contains('sentieon_tnhaplotyper2')) }
+            // TO-DO: sort out the following options which are options for mutect2
+            // ext.args         = { params.ignore_soft_clipped_bases ? "--dont-use-soft-clipped-bases true --f1r2-tar-gz ${task.ext.prefix}.f1r2.tar.gz" : "--f1r2-tar-gz ${task.ext.prefix}.f1r2.tar.gz" }
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/" },
+                pattern: "*{vcf.gz,vcf.gz.tbi,stats}",
+                saveAs: { meta.num_intervals > 1 ? null : "tnhaplotyper2/${meta.id}/${it}" }
+            ]
+        }
+
+        // PAIR_VARIANT_CALLING
+        withName: 'SENTIEON_TNHAPLOTYPER2_PAIRED' {
+            ext.args2       = { params.ignore_soft_clipped_bases ?
+                                "--trim_soft_clip --normal_sample ${meta.patient}_${meta.normal_id} --tumor_sample ${meta.patient}_${meta.tumor_id}" :
+                                "--normal_sample ${meta.patient}_${meta.normal_id} --tumor_sample ${meta.patient}_${meta.tumor_id}" }
+            ext.args3       = { "--tumor_sample ${meta.patient}_${meta.tumor_id}" }
+            ext.args4       = { "--normal_sample ${meta.patient}_${meta.normal_id} --tumor_sample ${meta.patient}_${meta.tumor_id}" }
+        }
+
+        withName: 'MERGE_TNHAPLOTYPER2.*' {
+            ext.prefix       = { "${meta.id}.tnhaplotyper2" }
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/tnhaplotyper2/${meta.id}" },
+                pattern: "*{vcf.gz,vcf.gz.tbi}"
+            ]
+        }
+
+        withName: 'SENTIEON_TNFILTER' {
+            ext.prefix = { "${meta.id}.filtered" }
+            ext.args2       = { "--normal_sample ${meta.patient}_${meta.normal_id} --tumor_sample ${meta.patient}_${meta.tumor_id}" }
+        }
+
+        // TO-DO: I guess that kind of prefixing with the name of the subworkflow is necessary since the user may request tnhaplotyper2 and mutect2 in the same run.
+        // TO-DO: Add a similar prefixing for the mutect2-subworkflow in mutect2.config `
+        withName: '.*_SENTIEON_TNHAPLOTYPER2:GATHERPILEUPSUMMARIES.*' {
+            ext.prefix       = { "${meta.id}.tnhaplotyper2" }
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/tnhaplotyper2/${meta.id}/" },
+                saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+            ]
+        }
+
+        withName: '.*_SENTIEON_TNHAPLOTYPER2:GETPILEUPSUMMARIES.*' {
+            ext.prefix       = { meta.num_intervals <= 1 ? "${meta.id}.tnhaplotyper2" : "${meta.id}.tnhaplotyper2.${intervals.simpleName}" }
+            publishDir        = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/" },
+                pattern: "*.table",
+                saveAs: { meta.num_intervals > 1 ? null : "tnhaplotyper2/${meta.id}/${it}" }
+            ]
+        }
+
+        withName: '.*_SENTIEON_TNHAPLOTYPER2:CALCULATECONTAMINATION' {
+            ext.prefix       = { "${meta.id}.tnhaplotyper2" }
+            ext.args         = { "-tumor-segmentation ${meta.id}.tnhaplotyper2.segmentation.table" }
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/tnhaplotyper2/${meta.id}" },
+                saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+            ]
+        }
+
+        if (params.joint_tnhaplotyper2) {
+            withName: '.*_SENTIEON_TNHAPLOTYPER2:CALCULATECONTAMINATION' {
+                publishDir       = [
+                    mode: params.publish_dir_mode,
+                    path: { "${params.outdir}/variant_calling/tnhaplotyper2/${meta.patient}" },
+                    saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+                ]
+            }
+        }
+
+        withName: '.*_SENTIEON_TNHAPLOTYPER2:SENTIEON_TNFILTER.*' {
+            ext.prefix       = {"${meta.id}.tnhaplotyper2.filtered"}
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/" },
+                saveAs: { filename -> filename.equals('versions.yml') ? null : "tnhaplotyper2/${meta.id}/${filename}" }
+            ]
+        }
+
+        // TO-DO: If possible, get rid of MERGEMUTECTSTATS in the tnhaplotyper2-subworkflow and use sentieon version instead.
+        withName: '.*_SENTIEON_TNHAPLOTYPER2:MERGEMUTECTSTATS' {
+            ext.prefix       = { "${meta.id}.tnhaplotyper2" }
+            publishDir       = [
+                mode: params.publish_dir_mode,
+                path: { "${params.outdir}/variant_calling/tnhaplotyper2/${meta.id}/" },
+                saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+            ]
+        }
+
+    }
+}

conf/test.config

@@ -76,6 +76,16 @@ process {
         ext.args = { "--f1r2-tar-gz ${task.ext.prefix}.f1r2.tar.gz --normal-sample normal" }
     }
 
+    withName: '.*:SENTIEON_TNHAPLOTYPER2.*'{
+        //sample name from when the test data was generated
+        ext.args2 = { "--normal_sample normal --tumor_sample tumour" }
+    }
+
+    withName: '.*:SENTIEON_TNFILTER.*' {
+        ext.prefix = { "${meta.id}.filtered" }
+        ext.args2 = { "--tumor_sample tumour --normal_sample normal" }
+    }
+
     withName: '.*:FILTERVARIANTTRANCHES'{
         ext.args = { "--info-key CNN_1D --indel-tranche 0" }
     }

conf/test/tools_somatic_tnhaplotyper2.config

@@ -0,0 +1,26 @@
+/*
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+    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 {
+    // TO-DO: Put test-data in repo.
+    // TO-DO: Figure out which of the following settings are relevant for tnhaplotyper2
+    input                  = "/home/ubuntu/test_data_tnhaplotyper2/fixed_readgroup.csv"
+    chr_dir                = params.test_data['homo_sapiens']['genome']['genome_21_chromosomes_dir']
+    dbsnp                  = params.test_data['homo_sapiens']['genome']['dbsnp_138_hg38_21_vcf_gz']
+    fasta                  = params.test_data['homo_sapiens']['genome']['genome_21_fasta']
+    germline_resource      = params.test_data['homo_sapiens']['genome']['gnomad_r2_1_1_21_vcf_gz']
+    intervals              = params.test_data['homo_sapiens']['genome']['genome_21_multi_interval_bed']
+    pon                    = params.test_data['homo_sapiens']['genome']['mills_and_1000g_indels_21_vcf_gz']
+    nucleotides_per_second = 20
+    step                   = 'variant_calling'
+    tools                  = 'sentieon_tnhaplotyper2'
+    wes                    = true
+}

modules.json

@@ -424,6 +424,16 @@
                         "git_sha": "e809c6b078d5343bdf8b5b2b78483096a2b5a973",
                         "installed_by": ["modules"]
                     },
+                    "sentieon/tnfilter": {
+                        "branch": "master",
+                        "git_sha": "388b763df527bb267cde359ded753a05e33fc0aa",
+                        "installed_by": ["modules"]
+                    },
+                    "sentieon/tnhaplotyper2": {
+                        "branch": "master",
+                        "git_sha": "30e382429b9761f39906982cc2cb1c31e0706ebe",
+                        "installed_by": ["modules"]
+                    },
                     "sentieon/varcal": {
                         "branch": "master",
                         "git_sha": "e809c6b078d5343bdf8b5b2b78483096a2b5a973",