Wdlupdate (#8)

* updated wdl to gatk4.0.5.0 release * realignment_index_bundle variable to json
gatk-workflows · Jun 15, 2018 · 0ab43b7 · 0ab43b7
1 parent ebb6dde
commit 0ab43b7
Show file tree

Hide file tree

Showing 6 changed files with 197 additions and 25 deletions.
diff --git a/mutect2-normal-normal.inputs.json b/mutect2-normal-normal.inputs.json
@@ -34,6 +34,6 @@
   "Mutect2NormalNormal.scatter_count": "10",
 
   "##_COMMENT4": "Docker",
-  "Mutect2NormalNormal.gatk_docker": "broadinstitute/gatk:4.0.4.0"
+  "Mutect2NormalNormal.gatk_docker": "broadinstitute/gatk:4.0.5.0"
 }
 
diff --git a/mutect2-normal-normal.wdl b/mutect2-normal-normal.wdl
@@ -32,13 +32,12 @@ workflow Mutect2NormalNormal {
 	Array[File] bams
 	Array[File] bais
 	File? pon
-	File? pon_index
 	File? gnomad
-	File? gnomad_index
 	File? variants_for_contamination
-    File? variants_for_contamination_index
 	Boolean? run_orientation_bias_filter
 	Array[String]? artifact_modes
+	File? realignment_index_bundle
+    String? realignment_extra_args
 	String? m2_extra_args
     String? m2_extra_filtering_args
     Boolean? make_bamout
@@ -68,15 +67,14 @@ workflow Mutect2NormalNormal {
                     normal_bam = normal_bam,
                     normal_bai = normal_bai,
                     pon = pon,
-                    pon_index = pon_index,
                     scatter_count = scatter_count,
                     gnomad = gnomad,
-                    gnomad_index = gnomad_index,
                     variants_for_contamination = variants_for_contamination,
-                    variants_for_contamination_index = variants_for_contamination_index,
                     run_orientation_bias_filter = run_orientation_bias_filter,
                     preemptible_attempts = preemptible_attempts,
                     artifact_modes = artifact_modes,
+                    realignment_index_bundle = realignment_index_bundle,
+                    realignment_extra_args = realignment_extra_args,
                     m2_extra_args = m2_extra_args,
                     m2_extra_filtering_args = m2_extra_filtering_args,
                     gatk_override = gatk_override,
@@ -155,6 +153,7 @@ task CountFalsePositives {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: "5 GB"
         disks: "local-disk " + 500 + " HDD"
     }

diff --git a/mutect2.exome.inputs.json b/mutect2.exome.inputs.json
@@ -1,7 +1,7 @@
 {
   "##_COMMENT1": "Runtime",
   "##Mutect2.oncotator_docker": "(optional) String?",
-  "Mutect2.gatk_docker": "broadinstitute/gatk:4.0.4.0",
+  "Mutect2.gatk_docker": "broadinstitute/gatk:4.0.5.0",
 
   "##_COMMENT2": "Workflow options",
   "Mutect2.intervals": "gs://gatk-best-practices/somatic-b37/whole_exome_agilent_1.1_refseq_plus_3_boosters.Homo_sapiens_assembly19.baits.interval_list",
@@ -28,6 +28,7 @@
   "Mutect2.gnomad_index": "gs://gatk-best-practices/somatic-b37/af-only-gnomad.raw.sites.vcf.idx",
   "Mutect2.variants_for_contamination": "gs://gatk-best-practices/somatic-b37/small_exac_common_3.vcf",
   "Mutect2.variants_for_contamination_index": "gs://gatk-best-practices/somatic-b37/small_exac_common_3.vcf.idx",
+  "##Mutect2.realignment_index_bundle": "File? (optional)",
 
   "##_COMMENT5": "Secondary resources",
   "Mutect2.onco_ds_tar_gz": "gs://gatk-best-practices/somatic-b37/oncotator_v1_ds_April052016.tar.gz",

diff --git a/mutect2.wdl b/mutect2.wdl
@@ -41,6 +41,7 @@
 ## onco_ds_tar_gz, default_config_file: Oncotator datasources and config file
 ## sequencing_center, sequence_source: metadata for Oncotator
 ## filter_oncotator_maf: Whether the MAF generated by oncotator should have the filtered variants removed. Default: true
+## realignment_index_bundle: resource for FilterAlignmentArtifacts, which runs if and only if it is specified.  Generated by BwaMemIndexImageCreator.
 ##
 ## Outputs :
 ## - One VCF file and its index with primary filtering applied; secondary filtering and functional annotation if requested; a bamout.bam
@@ -73,6 +74,8 @@ workflow Mutect2 {
     File? gnomad_index
     File? variants_for_contamination
     File? variants_for_contamination_index
+    File? realignment_index_bundle
+    String? realignment_extra_args
     Boolean? run_orientation_bias_filter
     Boolean run_ob_filter = select_first([run_orientation_bias_filter, false])
     Array[String]? artifact_modes
@@ -115,6 +118,8 @@ workflow Mutect2 {
     String oncotator_docker_or_default = select_first([oncotator_docker, "broadinstitute/oncotator:1.9.9.0"])
     Boolean? filter_oncotator_maf
     Boolean filter_oncotator_maf_or_default = select_first([filter_oncotator_maf, true])
+    Boolean? filter_funcotations
+    Boolean filter_funcotations_or_default = select_first([filter_funcotations, true])
     String? oncotator_extra_args
 
     Int? preemptible_attempts
@@ -302,8 +307,26 @@ workflow Mutect2 {
         }
     }
 
+    if (defined(realignment_index_bundle)) {
+        File realignment_filter_input = select_first([FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
+        File realignment_filter_input_idx = select_first([FilterByOrientationBias.filtered_vcf_index, Filter.filtered_vcf_index])
+        call FilterAlignmentArtifacts {
+            input:
+                gatk_override = gatk_override,
+                bam = tumor_bam,
+                bai = tumor_bai,
+                realignment_index_bundle = select_first([realignment_index_bundle]),
+                realignment_extra_args = realignment_extra_args,
+                gatk_docker = gatk_docker,
+                compress = compress,
+                output_name = filtered_name,
+                input_vcf = realignment_filter_input,
+                input_vcf_idx = realignment_filter_input_idx
+        }
+    }
+
     if (run_oncotator_or_default) {
-        File oncotate_vcf_input = select_first([FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
+        File oncotate_vcf_input = select_first([FilterAlignmentArtifacts.filtered_vcf, FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
         call oncotate_m2 {
             input:
                 m2_vcf = oncotate_vcf_input,
@@ -323,8 +346,8 @@ workflow Mutect2 {
     }
 
     if (run_funcotator_or_default) {
-        File funcotate_vcf_input = select_first([FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
-        File funcotate_vcf_input_index = select_first([FilterByOrientationBias.filtered_vcf_index, Filter.filtered_vcf_index])
+        File funcotate_vcf_input = select_first([FilterAlignmentArtifacts.filtered_vcf, FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
+        File funcotate_vcf_input_index = select_first([FilterAlignmentArtifacts.filtered_vcf_index, FilterByOrientationBias.filtered_vcf_index, Filter.filtered_vcf_index])
         call Funcotate {
             input:
                 m2_vcf = funcotate_vcf_input,
@@ -341,15 +364,16 @@ workflow Mutect2 {
                 annotation_defaults = annotation_defaults,
                 annotation_overrides = annotation_overrides,
                 gatk_docker = gatk_docker,
-                gatk_override = gatk_override
+                gatk_override = gatk_override,
+                filter_funcotations = filter_funcotations_or_default
         }
     }
 
     output {
         File unfiltered_vcf = MergeVCFs.merged_vcf
         File unfiltered_vcf_index = MergeVCFs.merged_vcf_index
-        File filtered_vcf = select_first([FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
-        File filtered_vcf_index = select_first([FilterByOrientationBias.filtered_vcf_index, Filter.filtered_vcf_index])
+        File filtered_vcf = select_first([FilterAlignmentArtifacts.filtered_vcf, FilterByOrientationBias.filtered_vcf, Filter.filtered_vcf])
+        File filtered_vcf_index = select_first([FilterAlignmentArtifacts.filtered_vcf_index, FilterByOrientationBias.filtered_vcf_index, Filter.filtered_vcf_index])
         File? contamination_table = CalculateContamination.contamination_table
 
         File? oncotated_m2_maf = oncotate_m2.oncotated_m2_maf
@@ -358,6 +382,7 @@ workflow Mutect2 {
         File? preadapter_detail_metrics = CollectSequencingArtifactMetrics.pre_adapter_metrics
         File? bamout = MergeBamOuts.merged_bam_out
         File? bamout_index = MergeBamOuts.merged_bam_out_index
+        File? maf_segments = CalculateContamination.maf_segments
     }
 }
 
@@ -400,6 +425,7 @@ task SplitIntervals {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -481,6 +507,7 @@ task M2 {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -529,6 +556,7 @@ task MergeVCFs {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -576,12 +604,13 @@ task MergeBamOuts {
 
         gatk --java-options "-Xmx${command_mem}m" SortSam -I unsorted.out.bam \
             -O ${output_vcf_name}.out.bam \
-            --SORT_ORDER coordinate
-        gatk --java-options "-Xmx${command_mem}m" BuildBamIndex -I ${output_vcf_name}.out.bam
+            --SORT_ORDER coordinate -VALIDATION_STRINGENCY LENIENT
+        gatk --java-options "-Xmx${command_mem}m" BuildBamIndex -I ${output_vcf_name}.out.bam -VALIDATION_STRINGENCY LENIENT
     >>>
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -624,6 +653,7 @@ task CollectSequencingArtifactMetrics {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -676,6 +706,7 @@ task CalculateContamination {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: command_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -729,6 +760,7 @@ task Filter {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -779,6 +811,61 @@ task FilterByOrientationBias {
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
+        memory: command_mem + " MB"
+        disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
+        preemptible: select_first([preemptible_attempts, 10])
+        cpu: select_first([cpu, 1])
+    }
+
+    output {
+        File filtered_vcf = "${output_vcf}"
+        File filtered_vcf_index = "${output_vcf_index}"
+    }
+}
+
+task FilterAlignmentArtifacts {
+    #input
+    File? gatk_override
+    File input_vcf
+    File input_vcf_idx
+    File bam
+    File bai
+    String output_name
+    Boolean compress
+    String output_vcf = output_name + if compress then ".vcf.gz" else ".vcf"
+    String output_vcf_index = output_vcf +  if compress then ".tbi" else ".idx"
+    File realignment_index_bundle
+    String? realignment_extra_args
+
+    # runtime
+    String gatk_docker
+    Int? mem
+    Int? preemptible_attempts
+    Int? disk_space
+    Int? cpu
+    Boolean use_ssd = false
+
+    # Mem is in units of GB but our command and memory runtime values are in MB
+    Int machine_mem = if defined(mem) then mem * 1000 else 9000
+    Int command_mem = machine_mem - 500
+
+    command {
+        set -e
+
+        export GATK_LOCAL_JAR=${default="/root/gatk.jar" gatk_override}
+
+        gatk --java-options "-Xmx${command_mem}m" FilterAlignmentArtifacts \
+            -V ${input_vcf} \
+            -I ${bam} \
+            --bwa-mem-index-image ${realignment_index_bundle} \
+            ${realignment_extra_args} \
+            -O ${output_vcf}
+    }
+
+    runtime {
+        docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: command_mem + " MB"
         disks: "local-disk " + select_first([disk_space, 100]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 10])
@@ -842,7 +929,7 @@ task oncotate_m2 {
         fi
 
         ${default="/root/oncotator_venv/bin/oncotator" oncotator_exe} --db-dir onco_dbdir/ -c $HOME/tx_exact_uniprot_matches.AKT1_CRLF2_FGFR1.txt  \
-            -v ${m2_vcf} ${case_id}.maf.annotated hg19 -i VCF -o TCGAMAF --skip-no-alt --infer-onps --collapse-number-annotations --log_name oncotator.log \
+            -v ${m2_vcf} ${case_id}.maf.annotated hg19 -i VCF -o TCGAMAF --skip-no-alt --collapse-number-annotations --log_name oncotator.log \
             -a Center:${default="Unknown" sequencing_center} \
             -a source:${default="Unknown" sequence_source} \
             -a normal_barcode:${control_id} \
@@ -906,12 +993,14 @@ task Funcotate {
     Array[String]? transcript_selection_list
     Array[String]? annotation_defaults
     Array[String]? annotation_overrides
+    Boolean filter_funcotations
 
     # ==============
     # Process input args:
     String transcript_selection_arg = if defined(transcript_selection_list) then " --transcript-list " else ""
     String annotation_def_arg = if defined(annotation_defaults) then " --annotation-default " else ""
     String annotation_over_arg = if defined(annotation_overrides) then " --annotation-override " else ""
+    String filter_funcotations_args = if (filter_funcotations) then " --remove-filtered-variants " else ""
     # ==============
 
     # runtime
@@ -962,11 +1051,13 @@ task Funcotate {
             ${"--transcript-selection-mode " + transcript_selection_mode} \
             ${transcript_selection_arg}${default="" sep=" --transcript-list " transcript_selection_list} \
             ${annotation_def_arg}${default="" sep=" --annotation-default " annotation_defaults} \
-            ${annotation_over_arg}${default="" sep=" --annotation-override " annotation_overrides}
+            ${annotation_over_arg}${default="" sep=" --annotation-override " annotation_overrides} \
+            ${filter_funcotations_args}
     >>>
 
     runtime {
         docker: gatk_docker
+        bootDiskSizeGb: 12
         memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, default_disk_space_gb]) + if use_ssd then " SSD" else " HDD"
         preemptible: select_first([preemptible_attempts, 3])