Standardized use of mem parameters in CNV WDLs.

broadinstitute · Jan 17, 2018 · f9b2701 · f9b2701
1 parent bec3701
commit f9b2701
Show file tree

Hide file tree

Showing 5 changed files with 31 additions and 34 deletions.
diff --git a/scripts/cnv_wdl/cnv_common_tasks.wdl b/scripts/cnv_wdl/cnv_common_tasks.wdl
@@ -92,6 +92,9 @@ task CollectCounts {
     Int? preemptible_attempts
     Int? disk_space_gb
 
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 1000
+
     # Sample name is derived from the bam filename
     String base_filename = basename(bam, ".bam")
     String counts_filename = if !defined(format) then "${base_filename}.counts.hdf5" else "${base_filename}.counts.tsv"
@@ -100,7 +103,7 @@ task CollectCounts {
         set -e
         export GATK_LOCAL_JAR=${default="/root/gatk.jar" gatk4_jar_override}
 
-        gatk --java-options "-Xmx${default="8" mem}g" CollectFragmentCounts \
+        gatk --java-options "-Xmx${command_mem}m" CollectFragmentCounts \
             --input ${bam} \
             -L ${intervals} \
             --format ${default="HDF5" format} \
@@ -110,7 +113,7 @@ task CollectCounts {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: select_first([mem, 8]) + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, ceil(size(bam, "GB")) + 50]) + " HDD"
         preemptible: select_first([preemptible_attempts, 5])
     }
@@ -137,8 +140,7 @@ task CollectAllelicCounts {
     Int? preemptible_attempts
     Int? disk_space_gb
 
-    # 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 13000
+    Int machine_mem = select_first([mem, 13]) * 1000
     Int command_mem = machine_mem - 1000
 
     # Sample name is derived from the bam filename

diff --git a/scripts/cnv_wdl/germline/cnv_germline_case_workflow.wdl b/scripts/cnv_wdl/germline/cnv_germline_case_workflow.wdl
@@ -202,8 +202,8 @@ task DetermineGermlineContigPloidyCaseMode {
 
     # We do not expose Hybrid ADVI parameters -- the default values are decent
 
-    Int machine_mem = if defined(mem) then select_first([mem]) else 8
-    Float command_mem = machine_mem - 0.5
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 500
 
     # If optional output_dir not specified, use "out"
     String output_dir_ = select_first([output_dir, "out"])
@@ -218,7 +218,7 @@ task DetermineGermlineContigPloidyCaseMode {
         mkdir input-contig-ploidy-model
         tar xzf ${contig_ploidy_model_tar} -C input-contig-ploidy-model
 
-        gatk --java-options "-Xmx${machine_mem}g" DetermineGermlineContigPloidy \
+        gatk --java-options "-Xmx${command_mem}m" DetermineGermlineContigPloidy \
             --input ${sep=" --input " read_count_files} \
             --model input-contig-ploidy-model \
             --output ${output_dir_} \
@@ -232,7 +232,7 @@ task DetermineGermlineContigPloidyCaseMode {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: command_mem + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, 150]) + " HDD"
         preemptible: select_first([preemptible_attempts, 2])
         cpu: select_first([cpu, 8])
@@ -293,8 +293,8 @@ task GermlineCNVCallerCaseMode {
     Float? caller_admixing_rate
     Boolean? disable_annealing
 
-    Int machine_mem = if defined(mem) then select_first([mem]) else 8
-    Float command_mem = machine_mem - 0.5
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 500
 
     # If optional output_dir not specified, use "out"
     String output_dir_ = select_first([output_dir, "out"])
@@ -312,7 +312,7 @@ task GermlineCNVCallerCaseMode {
         mkdir gcnv-model
         tar xzf ${gcnv_model_tar} -C gcnv-model
 
-        gatk --java-options "-Xmx${machine_mem}g"  GermlineCNVCaller \
+        gatk --java-options "-Xmx${command_mem}m"  GermlineCNVCaller \
             --run-mode CASE \
             --input ${sep=" --input " read_count_files} \
             --contig-ploidy-calls contig-ploidy-calls-dir \
@@ -353,7 +353,7 @@ task GermlineCNVCallerCaseMode {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: command_mem + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, 150]) + " HDD"
         preemptible: select_first([preemptible_attempts, 2])
         cpu: select_first([cpu, 8])

diff --git a/scripts/cnv_wdl/germline/cnv_germline_cohort_workflow.wdl b/scripts/cnv_wdl/germline/cnv_germline_cohort_workflow.wdl
@@ -248,8 +248,8 @@ task DetermineGermlineContigPloidyCohortMode {
 
     # We do not expose Hybrid ADVI parameters -- the default values are decent
 
-    Int machine_mem = if defined(mem) then select_first([mem]) else 8
-    Float command_mem = machine_mem - 0.5
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 500
 
     # If optional output_dir not specified, use "out"
     String output_dir_ = select_first([output_dir, "out"])
@@ -261,7 +261,7 @@ task DetermineGermlineContigPloidyCohortMode {
         export MKL_NUM_THREADS=${default=8 cpu}
         export OMP_NUM_THREADS=${default=8 cpu}
 
-        gatk --java-options "-Xmx${machine_mem}g"  DetermineGermlineContigPloidy \
+        gatk --java-options "-Xmx${command_mem}m"  DetermineGermlineContigPloidy \
             --input ${sep=" --input " read_count_files} \
             --contig-ploidy-priors ${contig_ploidy_priors} \
             --output ${output_dir_} \
@@ -278,7 +278,7 @@ task DetermineGermlineContigPloidyCohortMode {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: command_mem + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, 150]) + " HDD"
         preemptible: select_first([preemptible_attempts, 2])
         cpu: select_first([cpu, 8])
@@ -349,8 +349,8 @@ task GermlineCNVCallerCohortMode {
     Float? caller_admixing_rate
     Boolean? disable_annealing
 
-    Int machine_mem = if defined(mem) then select_first([mem]) else 8
-    Float command_mem = machine_mem - 0.5
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 500
 
     # If optional output_dir not specified, use "out"
     String output_dir_ = select_first([output_dir, "out"])
@@ -365,7 +365,7 @@ task GermlineCNVCallerCohortMode {
         mkdir contig-ploidy-calls-dir
         tar xzf ${contig_ploidy_calls_tar} -C contig-ploidy-calls-dir
 
-        gatk --java-options "-Xmx${machine_mem}g"  GermlineCNVCaller \
+        gatk --java-options "-Xmx${command_mem}m"  GermlineCNVCaller \
             --run-mode COHORT \
             -L ${intervals} \
             --input ${sep=" --input " read_count_files} \
@@ -418,7 +418,7 @@ task GermlineCNVCallerCohortMode {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: command_mem + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, 150]) + " HDD"
         preemptible: select_first([preemptible_attempts, 2])
         cpu: select_first([cpu, 8])

diff --git a/scripts/cnv_wdl/somatic/cnv_somatic_pair_workflow.wdl b/scripts/cnv_wdl/somatic/cnv_somatic_pair_workflow.wdl
@@ -309,8 +309,7 @@ task DenoiseReadCounts {
     Int? preemptible_attempts
     Int disk_space_gb
 
-    # 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 13000
+    Int machine_mem = select_first([mem, 13]) * 1000
     Int command_mem = machine_mem - 1000
 
     command <<<
@@ -371,8 +370,7 @@ task ModelSegments {
     Int? preemptible_attempts
     Int disk_space_gb
 
-    # 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 13000
+    Int machine_mem = select_first([mem, 13]) * 1000
     # ModelSegments seems to need at least 3GB of overhead to run
     Int command_mem = machine_mem - 3000
 
@@ -449,8 +447,7 @@ task CallCopyRatioSegments {
     Int? preemptible_attempts
     Int disk_space_gb
 
-    # 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 7000
+    Int machine_mem = select_first([mem, 7]) * 1000
     Int command_mem = machine_mem - 1000
 
     command <<<
@@ -492,8 +489,7 @@ task PlotDenoisedCopyRatios {
     Int? preemptible_attempts
     Int disk_space_gb
 
-    # 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 7000
+    Int machine_mem = select_first([mem, 7]) * 1000
     Int command_mem = machine_mem - 1000
 
     # If optional output_dir not specified, use "out"
@@ -546,8 +542,7 @@ task PlotModeledSegments {
     Int? preemptible_attempts
     Int disk_space_gb
 
-    # 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 7000
+    Int machine_mem = select_first([mem, 7]) * 1000
     Int command_mem = machine_mem - 1000
 
     # If optional output_dir not specified, use "out"

diff --git a/scripts/cnv_wdl/somatic/cnv_somatic_panel_workflow.wdl b/scripts/cnv_wdl/somatic/cnv_somatic_panel_workflow.wdl
@@ -105,14 +105,14 @@ task CreateReadCountPanelOfNormals {
     Int? preemptible_attempts
     Int? disk_space_gb
 
-    Int machine_mem = if defined(mem) then select_first([mem]) else 8
-    Float command_mem = machine_mem - 0.5
+    Int machine_mem = select_first([mem, 8]) * 1000
+    Int command_mem = machine_mem - 500
 
     command <<<
         set -e
         export GATK_LOCAL_JAR=${default="/root/gatk.jar" gatk4_jar_override}
 
-        gatk --java-options "-Xmx${machine_mem}g" CreateReadCountPanelOfNormals \
+        gatk --java-options "-Xmx${command_mem}m" CreateReadCountPanelOfNormals \
             --input ${sep=" --input " read_count_files} \
             --minimum-interval-median-percentile ${default="10.0" minimum_interval_median_percentile} \
             --maximum-zeros-in-sample-percentage ${default="5.0" maximum_zeros_in_sample_percentage} \
@@ -127,7 +127,7 @@ task CreateReadCountPanelOfNormals {
 
     runtime {
         docker: "${gatk_docker}"
-        memory: command_mem + " GB"
+        memory: machine_mem + " MB"
         disks: "local-disk " + select_first([disk_space_gb, 150]) + " HDD"
         preemptible: select_first([preemptible_attempts, 2])
     }