🔧 accept cram input

🔧 Svaba CRAM adjustments
kids-first · Sep 7, 2022 · a9de7f5 · a9de7f5
1 parent 2f92e5f
commit a9de7f5
Show file tree

Hide file tree

Showing 3 changed files with 44 additions and 36 deletions.
diff --git a/docs/GERMLINE_SV_README.md b/docs/GERMLINE_SV_README.md
@@ -6,8 +6,8 @@
 
 The Kids First Data Resource Center (KFDRC) Germline Structural Variant (SV)
 Caller Workflow is a common workflow language (CWL) implmentation to generate
-SV calls from an aligned reads BAM file. The workflow makes use of Manta and
-SvABA to call varaiants then annotates these variants using AnnotSV.
+SV calls from an aligned reads BAM or CRAM file. The workflow makes use of
+Manta and SvABA to call varaiants then annotates these variants using AnnotSV.
 
 ## Relevant Softwares and Versions
 
@@ -50,8 +50,8 @@ and has built-in support for case-control experiments (e.g. tumor/normal, or
 trios or quads). In case/control mode, any number of cases and controls (but
 min of 1 case) can be input, and will jointly assemble all sequences together.
 If both a case and control are present, variants are output separately in
-"somatic" and "germline" VCFs. If only a single BAM is present (input with the
--t flag), a single SV and a single indel VCF will be emitted.
+"somatic" and "germline" VCFs. If only a single BAM/CRAM is present (input with
+the -t flag), a single SV and a single indel VCF will be emitted.
 
 A BWA-MEM index reference genome must also be supplied with -G.
 
@@ -65,10 +65,10 @@ potential pathogenicity and ii) filter out SV potential false positives.
 ## Input Files
 
 At the moment the workflow uses only a few inputs:
-- `germline_bam`: The germline BAM input that has been aligned to a reference
-  genome.
+- `germline_reads`: The germline BAM/CRAM input that has been aligned to a
+  reference genome.
 - `indexed_reference_fasta`: The reference genome fasta (and associated
-  indicies) to which the germline BAM was aligned.
+  indicies) to which the germline BAM/CRAM was aligned.
 - `annotsv_annotations_dir`: These annotations are simply those from the
   install-human-annotation installation process run during AnnotSV installation
 (see: https://github.com/lgmgeo/AnnotSV/#quick-installation). Specifically

diff --git a/tools/svaba.cwl b/tools/svaba.cwl
@@ -19,7 +19,7 @@ doc: |
   trios or quads). In case/control mode, any number of cases and controls (but
   min of 1 case) can be input, and will jointly assemble all sequences together.
   If both a case and control are present, variants are output separately in
-  "somatic" and "germline" VCFs. If only a single BAM is present (input with the
+  "somatic" and "germline" VCFs. If only a single SAM is present (input with the
   -t flag), a single SV and a single indel VCF will be emitted.
 
   A BWA-MEM index reference genome must also be supplied with -G.
@@ -30,9 +30,15 @@ requirements:
   - class: ResourceRequirement
     ramMin: $(inputs.ram * 1000)
     coresMin: $(inputs.cores)
-baseCommand: ['svaba','run']
+baseCommand: []
 arguments:
   - position: 0
+    shellQuote: false
+    valueFrom: >
+      seq_cache_populate.pl -r $PWD/ref_cache $(inputs.reference_genome.path)
+      && export REF_CACHE=$PWD/ref_cache/%2s/%2s/%s
+      && svaba run
+  - position: 11
     valueFrom: "--g-zip"
 inputs:
   tumor_bams:
@@ -41,9 +47,9 @@ inputs:
       items: File
       inputBinding:
         prefix: '--tumor-bam'
-    secondaryFiles: [{pattern: '^.bai', required: true}]
+    secondaryFiles: [{pattern: '^.bai', required: false}, {pattern: '.bai', required: false}, {pattern: '^.crai', required: false}, {pattern: '.crai', required: false}]
     inputBinding:
-      position: 0
+      position: 11
     doc: "Case BAM/CRAM/SAM file (eg tumor). Can input multiple."
   normal_bams:
     type:
@@ -52,26 +58,26 @@ inputs:
         items: File
         inputBinding:
           prefix: '--normal-bam'
-    secondaryFiles: [{pattern: '^.bai', required: true}]
+    secondaryFiles: [{pattern: '^.bai', required: false}, {pattern: '.bai', required: false}, {pattern: '^.crai', required: false}, {pattern: '.crai', required: false}]
     inputBinding:
-      position: 0
+      position: 11
     doc: "Control BAM/CRAM/SAM file (eg normal). Can input multiple. Optional."
-  reference_genome: { type: 'File', secondaryFiles: [{pattern: '.fai', required: true}, {pattern: '.64.amb', required: true}, {pattern: '.64.ann', required: true}, {pattern: '.64.bwt', required: true}, {pattern: '.64.pac', required: true}, {pattern: '.64.sa', required: true}], inputBinding: { prefix: '--reference-genome', position: 0 }, doc: "Path to indexed reference genome to be used by BWA-MEM." }
+  reference_genome: { type: 'File', secondaryFiles: [{pattern: '.fai', required: true}, {pattern: '.64.amb', required: true}, {pattern: '.64.ann', required: true}, {pattern: '.64.bwt', required: true}, {pattern: '.64.pac', required: true}, {pattern: '.64.sa', required: true}], inputBinding: { prefix: '--reference-genome', position: 11 }, doc: "Path to indexed reference genome to be used by BWA-MEM." }
 
-  dbsnp_vcf: { type: 'File?', inputBinding: { prefix: '--dbsnp-vcf', position: 0 }, doc: "DBsnp database (VCF) to compare indels against" }
-  region_file: { type: 'File?', inputBinding: { prefix: '--region-file', position: 0 }, doc: "Run on targeted intervals. Accepts BED file" }
-  blacklist: { type: 'File?', inputBinding: { prefix: '--blacklist', position: 0 }, doc: "BED-file with blacklisted regions to not extract any reads from." }
-  germline_sv_database: { type: 'File?', inputBinding: { prefix: '--germline-sv-database', position: 0 }, doc: "BED file containing sites of known germline SVs. Used as additional filter for somatic SV detection." }
+  dbsnp_vcf: { type: 'File?', inputBinding: { prefix: '--dbsnp-vcf', position: 11 }, doc: "DBsnp database (VCF) to compare indels against" }
+  region_file: { type: 'File?', inputBinding: { prefix: '--region-file', position: 11 }, doc: "Run on targeted intervals. Accepts BED file" }
+  blacklist: { type: 'File?', inputBinding: { prefix: '--blacklist', position: 11 }, doc: "BED-file with blacklisted regions to not extract any reads from." }
+  germline_sv_database: { type: 'File?', inputBinding: { prefix: '--germline-sv-database', position: 11 }, doc: "BED file containing sites of known germline SVs. Used as additional filter for somatic SV detection." }
 
-  germline: { type: 'boolean?', inputBinding: { prefix: '--germline', position: 0}, doc: "Sets recommended settings for case-only analysis (eg germline). (-I, -L5, assembles NM >= 3 reads)" }
-  rules: { type: 'boolean?', inputBinding: { valueFrom: "$(self ? '--rules all' : '')", position: 0 }, doc: "Default behavior is just assemble clipped/discordant/unmapped/gapped reads. Override?" }
-  highly_parallel: { type: 'boolean?', inputBinding: { prefix: '--hp', position: 0 }, doc: "Highly parallel. Don't write output until completely done. More memory, but avoids all thread-locks." }
-  no_interchrom_lookup: { type: 'boolean?', inputBinding: { prefix: '--no-interchrom-lookup', position: 0 }, doc: "Set true to not do mate-region lookup if mates are mapped to different chromosome." }
+  germline: { type: 'boolean?', inputBinding: { prefix: '--germline', position: 11}, doc: "Sets recommended settings for case-only analysis (eg germline). (-I, -L5, assembles NM >= 3 reads)" }
+  rules: { type: 'boolean?', inputBinding: { valueFrom: "$(self ? '--rules all' : '')", position: 11 }, doc: "Default behavior is just assemble clipped/discordant/unmapped/gapped reads. Override?" }
+  highly_parallel: { type: 'boolean?', inputBinding: { prefix: '--hp', position: 11 }, doc: "Highly parallel. Don't write output until completely done. More memory, but avoids all thread-locks." }
+  no_interchrom_lookup: { type: 'boolean?', inputBinding: { prefix: '--no-interchrom-lookup', position: 11 }, doc: "Set true to not do mate-region lookup if mates are mapped to different chromosome." }
 
-  mate_lookup_min: { type: 'int?', inputBinding: { prefix: '--mate-lookup-min', position: 0 }, doc: "Minimum number of somatic reads required to attempt mate-region lookup" }
+  mate_lookup_min: { type: 'int?', inputBinding: { prefix: '--mate-lookup-min', position: 11 }, doc: "Minimum number of somatic reads required to attempt mate-region lookup" }
 
-  output_basename: { type: 'string', inputBinding: { prefix: '--id-string', position: 0 }, doc: "String specifying the analysis ID to be used as part of ID common." }
-  cores: { type: 'int?', default: 16, inputBinding: { prefix: '--threads', position: 0 }, doc: "Use NUM threads to run svaba." }
+  output_basename: { type: 'string', inputBinding: { prefix: '--id-string', position: 11 }, doc: "String specifying the analysis ID to be used as part of ID common." }
+  cores: { type: 'int?', default: 16, inputBinding: { prefix: '--threads', position: 11 }, doc: "Use NUM threads to run svaba." }
   ram: { type: 'int?', default: 16, doc: "Minimum ram to allocate to the task." }
 outputs:
   alignments: { type: 'File', outputBinding: { glob: "*.alignments.txt.gz" }, doc: "An ASCII plot of variant-supporting contigs and the BWA-MEM alignment of reads to the contigs" }

diff --git a/workflows/kfdrc-germline-sv-wf.cwl b/workflows/kfdrc-germline-sv-wf.cwl
@@ -11,8 +11,8 @@ doc: |
 
   The Kids First Data Resource Center (KFDRC) Germline Structural Variant (SV)
   Caller Workflow is a common workflow language (CWL) implmentation to generate
-  SV calls from an aligned reads BAM file. The workflow makes use of Manta and
-  SvABA to call varaiants then annotates these variants using AnnotSV.
+  SV calls from an aligned reads BAM or CRAM file. The workflow makes use of
+  Manta and SvABA to call varaiants then annotates these variants using AnnotSV.
 
   ## Relevant Softwares and Versions
 
@@ -55,8 +55,8 @@ doc: |
   trios or quads). In case/control mode, any number of cases and controls (but
   min of 1 case) can be input, and will jointly assemble all sequences together.
   If both a case and control are present, variants are output separately in
-  "somatic" and "germline" VCFs. If only a single BAM is present (input with the
-  -t flag), a single SV and a single indel VCF will be emitted.
+  "somatic" and "germline" VCFs. If only a single BAM/CRAM is present (input with
+  the -t flag), a single SV and a single indel VCF will be emitted.
 
   A BWA-MEM index reference genome must also be supplied with -G.
 
@@ -70,10 +70,10 @@ doc: |
   ## Input Files
 
   At the moment the workflow uses only a few inputs:
-  - `germline_bam`: The germline BAM input that has been aligned to a reference
-    genome.
+  - `germline_reads`: The germline BAM/CRAM input that has been aligned to a
+    reference genome.
   - `indexed_reference_fasta`: The reference genome fasta (and associated
-    indicies) to which the germline BAM was aligned.
+    indicies) to which the germline BAM/CRAM was aligned.
   - `annotsv_annotations_dir`: These annotations are simply those from the
     install-human-annotation installation process run during AnnotSV installation
   (see: https://github.com/lgmgeo/AnnotSV/#quick-installation). Specifically
@@ -136,8 +136,10 @@ inputs:
         {class: File, path: 6063901d357c3a53540ca81e, name: Homo_sapiens_assembly38.fasta.64.bwt},
         {class: File, path: 6063901c357c3a53540ca801, name: Homo_sapiens_assembly38.fasta.64.pac},
         {class: File, path: 60639015357c3a53540ca7a9, name: Homo_sapiens_assembly38.fasta.64.sa}]}
-  germline_bam: {type: 'File', secondaryFiles: [{pattern: '^.bai', required: false},
-      {pattern: '.bai', required: false}], doc: "Input BAM file", "sbg:fileTypes": "BAM"}
+  germline_reads: {type: 'File', secondaryFiles: [{pattern: '^.bai', required: false},
+      {pattern: '.bai', required: false}, {pattern: '^.crai', required: false}, {
+        pattern: '.crai', required: false}], doc: "Input BAM file", "sbg:fileTypes": "BAM,\
+      \ CRAM"}
 
   annotsv_annotations_dir: {type: 'File', doc: "TAR.GZ'd Directory containing AnnotSV\
       \ annotations", "sbg:fileTypes": "TAR, TAR.GZ, TGZ", "sbg:suggestedValue": {
@@ -178,7 +180,7 @@ steps:
     run: ../tools/svaba.cwl
     in:
       tumor_bams:
-        source: germline_bam
+        source: germline_reads
         valueFrom: $([self])
       reference_genome: indexed_reference_fasta
       germline:
@@ -197,7 +199,7 @@ steps:
     in:
       reference: indexed_reference_fasta
       input_normal_reads:
-        source: germline_bam
+        source: germline_reads
         valueFrom: $([self])
       output_basename:
         source: output_basename