🧹 v1.0 release prep

README.md

@@ -1,9 +1,27 @@
 # Kids First DRC Tumor Only Pipeline
 
-This repository contains tools and workflows for processing of tumor-only samples.
-It is currently in beta phase.
-Much of the components have been borrowed from the Kids First Somatic Workflow.
-It can also be used to process PDX data by first pre-processing reads using the Xenome tool, explained more here in documentation.
+This repository contains tools and workflows for processing of tumor-only
+samples. The Kids First DRC recommends running the tumor only pipeline ONLY
+when no matched normal sample is available. If your data has matched normals
+we recommend running the [Kids First DRC Somatic Variant
+Workflow](https://github.com/kids-first/kf-somatic-workflow) instead. This
+workflow is not a traditional production pipeline run on all data, but rather
+is run at the user's request.
+
+When comparing the SNV outputs of this workflow to those of the somatic workflow,
+we have found the outputs to be considerably more noisy. To cut down on this
+noise, we have included some recommended inputs, parameters, and filters for
+Mutect2 [in our docs](./docs/MUTECT2_TUMOR_ONLY_FILTERING.md). In short we recommend:
+- Restrict the callable regions with a blacklist and Panel of Normals (PON)
+- Remove low support reads:
+   - Allele Depth (AD) == 0: WGS uninformative reads
+   - Variant Allele Frequency (VAF) < 1%: WXS noise
+- Remove potential germline variants: gnomAD AF > 0.00003
+- Only keep variants that are PASS
+- Rescue any variants that fall in hotspot regions/genes
+
+It can also be used to process PDX data by first pre-processing reads using the
+Xenome tool, explained more here in documentation.
 
 <p align="center">
   <img src="docs/kids_first_logo.svg" alt="Kids First repository logo" width="660px" />

subworkflows/kfdrc_controlfreec_sub_wf.cwl

@@ -24,6 +24,7 @@ inputs:
   b_allele: {type: ['null', File], doc: "germline calls, needed for BAF.  VarDict input recommended.  Tool will prefilter for germline and pass if expression given"}
   coeff_var: {type: float, default: 0.05, doc: "Coefficient of variantion to set window size.  Default 0.05 recommended"}
   cfree_sex: {type: ['null', {type: enum, name: sex, symbols: ["XX", "XY"] }], doc: "If known, XX for female, XY for male"}
+  tool_name: { type: 'string?', doc: "Tool name to use in outputs." }
 
 outputs:
   ctrlfreec_cnvs: {type: File, outputSource: rename_outputs/ctrlfreec_cnvs}
@@ -82,6 +83,7 @@ steps:
     in:
       input_files: [control_free_c/cnvs, control_free_c/cnvs_pvalue, control_free_c/config_script, control_free_c/ratio, control_free_c/sample_BAF, control_free_c/info_txt]
       input_pngs: control_free_c/pngs
+      tool_name: tool_name
       output_basename: output_basename
     out: [ctrlfreec_cnvs, ctrlfreec_pval, ctrlfreec_config, ctrlfreec_pngs, ctrlfreec_bam_ratio, ctrlfreec_baf, ctrlfreec_info]
 
@@ -95,4 +97,5 @@ steps:
       ctrlfreec_ratio: control_free_c/ratio
       sample_name: input_tumor_name
       output_basename: output_basename
+      tool_name: tool_name
     out: [ctrlfreec_ratio2seg]

subworkflows/kfdrc_manta_sub_wf.cwl

@@ -19,6 +19,7 @@ inputs:
   manta_cores: {type: "int?"}
   select_vars_mode: {type: ['null', {type: enum, name: select_vars_mode, symbols: ["gatk", "grep"]}], doc: "Choose 'gatk' for SelectVariants tool, or 'grep' for grep expression", default: "gatk"}
   annotsv_annotations_dir_tgz: {type: 'File', doc: "TAR.GZ'd Directory containing annotations for AnnotSV"}
+  tool_name: { type: 'string?', default: "manta", doc: "Tool name to use in outputs." }
 
 outputs:
   manta_prepass_vcf: {type: File, outputSource: rename_manta_samples/reheadered_vcf}
@@ -37,6 +38,7 @@ steps:
       cores: manta_cores
       reference: indexed_reference_fasta
       hg38_strelka_bed: hg38_strelka_bed
+      tool_name: tool_name
     out: [output_sv, small_indels]
 
   rename_manta_samples:
@@ -48,12 +50,10 @@ steps:
 
   gatk_selectvariants_manta:
     run: ../tools/gatk_selectvariants.cwl
-    label: GATK Select Manta PASS
     in:
       input_vcf: rename_manta_samples/reheadered_vcf
       output_basename: output_basename
-      tool_name:
-        valueFrom: $("manta")
+      tool_name: tool_name
       mode: select_vars_mode
     out: [pass_vcf]
 

subworkflows/kfdrc_mutect2_sub_wf.cwl

@@ -148,7 +148,9 @@ steps:
       reference: indexed_reference_fasta
       input_bams: mutect2/mutect2_bam
       enable_tool: make_bamout
-      output_basename: output_basename
+      output_basename:
+        source: [output_basename, tool_name]
+        valueFrom: $(self.join("."))
     out: [output]
 
   gatk_learn_orientation_bias:

tools/manta.cwl

@@ -2,46 +2,52 @@ cwlVersion: v1.0
 class: CommandLineTool
 id: kfdrc-manta-sv
 label: Manta sv caller
-doc: 'Calls structural variants.  Tool designed to pick correct run mode based on if tumor, normal, or both crams are given'
+doc: 'Calls structural variants. Tool designed to pick correct run mode based on if tumor, normal, or both crams are given'
 requirements:
   - class: ShellCommandRequirement
   - class: InlineJavascriptRequirement
   - class: ResourceRequirement
-    ramMin: ${ return inputs.ram * 1000 }
+    ramMin: $(inputs.ram * 1000)
     coresMin: $(inputs.cores)
   - class: DockerRequirement
     dockerPull: 'pgc-images.sbgenomics.com/d3b-bixu/manta:1.4.0'
 
-baseCommand: [/manta-1.4.0.centos6_x86_64/bin/configManta.py]
+baseCommand: []
 arguments:
-  - position: 1
+  - position: 0
     shellQuote: false
     valueFrom: >-
-      ${
-        var std = " --ref " + inputs.reference.path + " --callRegions " + inputs.hg38_strelka_bed.path + " --runDir=./ && ./runWorkflow.py -m local -j " + inputs.cores + " ";
-        var mv = " && mv results/variants/";
-        if (typeof inputs.input_tumor_aligned === 'undefined' || inputs.input_tumor_aligned === null){
-          var mv_cmd = mv + "diploidSV.vcf.gz " +  inputs.output_basename + ".manta.diploidSV.vcf.gz" + mv + "diploidSV.vcf.gz.tbi " + inputs.output_basename + ".manta.diploidSV.vcf.gz.tbi" + mv + "candidateSmallIndels.vcf.gz " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz" + mv + "candidateSmallIndels.vcf.gz.tbi " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz.tbi";
-          return "--bam ".concat(inputs.input_normal_aligned.path, std, mv_cmd);
-        }
-        else if (typeof inputs.input_normal_aligned === 'undefined' || inputs.input_normal_aligned === null){
-          var mv_cmd = mv + "tumorSV.vcf.gz " + inputs.output_basename + ".manta.tumorSV.vcf.gz" + mv + "tumorSV.vcf.gz.tbi " + inputs.output_basename + ".manta.tumorSV.vcf.gz.tbi" + mv + "candidateSmallIndels.vcf.gz " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz" + mv + "candidateSmallIndels.vcf.gz.tbi " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz.tbi";
-          return "--tumorBam " + inputs.input_tumor_aligned.path + std + mv_cmd;
-        }
-        else{
-          var mv_cmd = mv + "somaticSV.vcf.gz " + inputs.output_basename + ".manta.somaticSV.vcf.gz" + mv + "somaticSV.vcf.gz.tbi " + inputs.output_basename + ".manta.somaticSV.vcf.gz.tbi" + mv + "candidateSmallIndels.vcf.gz " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz" + mv + "candidateSmallIndels.vcf.gz.tbi " + inputs.output_basename + ".manta.candidateSmallIndels.vcf.gz.tbi";
-          return "--tumorBam " + inputs.input_tumor_aligned.path + " --normalBam " + inputs.input_normal_aligned.path + std + mv_cmd;
-        }
-      }
+      /manta-1.4.0.centos6_x86_64/bin/configManta.py --runDir=./
+  - position: 9
+    shellQuote: false
+    valueFrom: >-
+      $(inputs.input_normal_aligned != null ? inputs.input_tumor_aligned != null ? "--normalBam " + inputs.input_normal_aligned.path : "--bam " + inputs.input_normal_aligned.path : "")
+  - position: 10
+    shellQuote: false
+    prefix: "&&"
+    valueFrom: >-
+      ./runWorkflow.py -m local
+  - position: 20
+    shellQuote: false
+    valueFrom: >-
+      && mv results/variants/diploidSV.vcf.gz $([inputs.output_basename, inputs.tool_name, "diploidSV.vcf.gz"].join(".")) || :
+      && mv results/variants/diploidSV.vcf.gz.tbi $([inputs.output_basename, inputs.tool_name, "diploidSV.vcf.gz.tbi"].join(".")) || :
+      && mv results/variants/tumorSV.vcf.gz $([inputs.output_basename, inputs.tool_name, "tumorSV.vcf.gz"].join(".")) || :
+      && mv results/variants/tumorSV.vcf.gz.tbi $([inputs.output_basename, inputs.tool_name, "tumorSV.vcf.gz.tbi"].join(".")) || :
+      && mv results/variants/somaticSV.vcf.gz $([inputs.output_basename, inputs.tool_name, "somaticSV.vcf.gz"].join(".")) || :
+      && mv results/variants/somaticSV.vcf.gz.tbi $([inputs.output_basename, inputs.tool_name, "somaticSV.vcf.gz.tbi"].join(".")) || :
+      && mv results/variants/candidateSmallIndels.vcf.gz $([inputs.output_basename, inputs.tool_name, "candidateSmallIndels.vcf.gz"].join("."))
+      && mv results/variants/candidateSmallIndels.vcf.gz.tbi $([inputs.output_basename, inputs.tool_name, "candidateSmallIndels.vcf.gz.tbi"].join("."))
 
 inputs:
-    reference: {type: File, secondaryFiles: [^.dict, .fai]}
-    hg38_strelka_bed: {type: File, secondaryFiles: [.tbi]}
-    input_tumor_aligned: { type: 'File?', secondaryFiles: ["^.bai?", ".bai?", "^.crai?", ".crai?"], doc: "tumor BAM or CRAM" }
+    reference: {type: File, secondaryFiles: [^.dict, .fai], inputBinding: {position: 2, prefix: "--ref"}}
+    hg38_strelka_bed: {type: File, secondaryFiles: [.tbi], inputBinding: {position: 2, prefix: "--callRegions"}}
+    input_tumor_aligned: { type: 'File?', secondaryFiles: ["^.bai?", ".bai?", "^.crai?", ".crai?"], inputBinding: {position: 9, prefix: "--tumorBam"},  doc: "tumor BAM or CRAM" }
     input_normal_aligned: { type: 'File?', secondaryFiles: ["^.bai?", ".bai?", "^.crai?", ".crai?"], doc: "normal BAM or CRAM" }
-    cores: {type: ['null', int], default: 16}
+    cores: {type: ['null', int], default: 16, inputBinding: {position: 12, prefix: "-j"}}
     ram: {type: "int?", default: 10, doc: "GB of RAM an instance must have to run the task"}
     output_basename: string
+    tool_name: { type: 'string?', default: "manta", doc: "Tool name to use in outputs." }
 outputs:
   output_sv:
     type: File

tools/ubuntu_ratio2seg.cwl

@@ -29,7 +29,7 @@ arguments:
         smp = "$(inputs.sample_name)"
 
         ratio_file = open("$(inputs.ctrlfreec_ratio.path)")
-        out = open("$(inputs.output_basename).controlfreec.seg", "w")
+        out = open("$(inputs.output_basename).$(inputs.tool_name).seg", "w")
         out.write("ID\tchrom\tloc.start\tloc.end\tnum.mark\tseg.mean\n") 
         head = next(ratio_file)
         count = 0
@@ -70,6 +70,7 @@ inputs:
   ctrlfreec_ratio: File
   sample_name: string
   output_basename: string
+  tool_name: { type: 'string?', default: "controlfreec", doc: "Tool name to use for output filename" }
 
 outputs:
   ctrlfreec_ratio2seg:

tools/ubuntu_rename_outputs.cwl

@@ -25,9 +25,9 @@ arguments:
               parts.shift();
               var check = fname.substr(fname.length - 10);
               if (check == "config.txt") {
-                  cmd += "cp " + inputs.input_files[i].path + " " + inputs.output_basename + ".controlfreec.config.txt;";
+                  cmd += "cp " + inputs.input_files[i].path + " " + [inputs.output_basename, inputs.tool_name, "config.txt;"].join(".");
               } else {
-              fname = inputs.output_basename + ".controlfreec." + parts.join(".");
+              fname = [inputs.output_basename, inputs.tool_name, parts.join(".")].join(".");
               cmd += " cp " + inputs.input_files[i].path + " " + fname + ";";
               }
             }
@@ -41,7 +41,7 @@ arguments:
             fname = fname.replace(".txt", "");
             var parts = fname.split(".");
             parts.shift();
-            fname = inputs.output_basename + ".controlfreec." + parts.join(".") + ".png";
+            fname = [inputs.output_basename, inputs.tool_name, parts.join("."), "png"].join(".");
             cmd += " cp " + inputs.input_pngs[j].path + " " + fname + ";";
             }
           return cmd;
@@ -50,6 +50,7 @@ arguments:
 inputs:
   input_files: File[]
   input_pngs: File[]
+  tool_name: { type: 'string?', default: "controlfreec", doc: "Tool name to use in outputs." }
   output_basename: string
 outputs:
   ctrlfreec_baf: