🧠 memory hook for dedup

docs/KFDRC_SENTIEON_ALIGNMENT_GVCF_WORKFLOW_README.md

@@ -90,6 +90,12 @@ Metrics collection and contamination estimation are unchanged.
 | Gather VCFs                | Picard MergeVcfs                    | No splitting occurs in Sentieon     |
 | Metrics                    | Picard CollectVariantCallingMetrics | Picard CollectVariantCallingMetrics |
 
+### Workflow Troubleshooting
+
+- Sentieon tools scale up RAM usage to match allocated CPUs. If a task that is
+  running into memory issues, that can be solved by EITHER scaling UP the
+  task's allocated RAM and scaling DOWN the tasks allocated CPUs.
+
 ## Basic Info
 - [D3b dockerfiles](https://github.com/d3b-center/bixtools)
 - Testing Tools:

tools/sentieon_ReadWriter.cwl

@@ -12,10 +12,8 @@ requirements:
 - class: ShellCommandRequirement
 - class: InlineJavascriptRequirement
 - class: ResourceRequirement
-  coresMin: |
-    $(inputs.cpu_per_job ? inputs.cpu_per_job : 16)
-  ramMin: |
-    $(inputs.mem_per_job ? inputs.mem_per_job : 16000)
+  coresMin: $(inputs.cpu_per_job)
+  ramMin: $(inputs.mem_per_job * 1000)
 - class: DockerRequirement
   dockerPull: pgc-images.sbgenomics.com/hdchen/sentieon:202112.01_hifi
 - class: EnvVarRequirement
@@ -114,10 +112,12 @@ inputs:
   label: CPU per job
   doc: CPU per job
   type: int?
+  default: 16
 - id: mem_per_job
   label: Memory per job
-  doc: Memory per job[MB].
+  doc: Memory per job[GB].
   type: int?
+  default: 16
 
 outputs:
 - id: output_reads

tools/sentieon_dedup.cwl

@@ -17,28 +17,8 @@ doc: |-
 requirements:
 - class: ShellCommandRequirement
 - class: ResourceRequirement
-  coresMin: |-
-    ${
-        if (inputs.cpu_per_job)
-        {
-            return inputs.cpu_per_job
-        }
-        else
-        {
-            return 32
-        }
-    }
-  ramMin: |-
-    ${
-        if (inputs.mem_per_job)
-        {
-            return inputs.mem_per_job
-        }
-        else
-        {
-            return 32000
-        }
-    }
+  coresMin: $(inputs.cpu_per_job)
+  ramMin: $(inputs.mem_per_job * 1000)
 - class: DockerRequirement
   dockerPull: pgc-images.sbgenomics.com/hdchen/sentieon:202112.01_hifi
 - class: EnvVarRequirement
@@ -108,6 +88,12 @@ inputs:
   label: Basename for output files
   doc: Basename for the output files that are to be written.
   type: string?
+- id: cpu_per_job
+  type: int?
+  default: 32
+- id: mem_per_job
+  type: int?
+  default: 32
 
 outputs:
 - id: metrics_file

workflows/kfdrc_sentieon_alignment_wf.cwl

@@ -176,6 +176,10 @@ inputs:
   bamtofastq_ram: {type: 'int?', default: 2, doc: "RAM in GB to allocate to bamtofastq"}
   bwa_cpu: {type: 'int?', default: 36, doc: "CPUs to allocate to Sentieon BWA"}
   bwa_ram: {type: 'int?', default: 72, doc: "RAM in GB to allocate to Sentieon BWA"}
+  dedup_cpu: {type: 'int?', default: 32, doc: "CPUs to allocate to Sentieon DeDup"}
+  dedup_ram: {type: 'int?', default: 32, doc: "RAM in GB to allocate to Sentieon DeDup"}
+  bam_to_cram_cpu: {type: 'int?', default: 16, doc:  "CPUs to allocate to Sentieon BAM to CRAM"}
+  bam_to_cram_ram: {type: 'int?', default: 16, doc:  "RAM in GB to allocate to Sentieon BAM to CRAM"}
   run_t1k: {type: 'boolean?', default: true, doc: "Set to false to disable T1k HLA typing"}
   hla_dna_ref_seqs: {type: 'File?', doc: "FASTA file containing the HLA allele reference sequences for DNA.", "sbg:suggestedValue": {
       class: File, path: 6669ac8127374715fc3ba3c4, name: hla_v3.43.0_gencode_v39_dna_seq.fa}}
@@ -343,6 +347,8 @@ steps:
       sentieon_license: sentieon_license
       reference: untar_reference/indexed_fasta
       in_alignments: sentieon_bwa_mem_payloads/realgn_bam
+      cpu_per_job: dedup_cpu
+      mem_per_job: dedup_ram
     out: [metrics_file, out_alignments]
   sentieon_bqsr:
     run: ../tools/sentieon_bqsr.cwl
@@ -385,6 +391,8 @@ steps:
         valueFrom: $(self.nameroot).cram
       rm_cram_bai:
         valueFrom: $(1 == 1)
+      cpu_per_job: bam_to_cram_cpu
+      mem_per_job: bam_to_cram_ram
     out: [output_reads]
   sentieon_hsmetrics:
     run: ../tools/sentieon_HsMetricAlgo.cwl