Merge pull request #8 from biowdl/run_time

Run time settings and additional adjustments

biopet.wdl

@@ -32,11 +32,15 @@ task ScatterRegions {
     Int? scatterSize
     File? regions
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
         mkdir -p ${outputDirPath}
-        java -Xmx2G -jar ${tool_jar} \
+        java -Xmx${mem}G -jar ${tool_jar} \
           -R ${ref_fasta} \
           -o ${outputDirPath} \
           ${"-s " + scatterSize} \
@@ -46,6 +50,10 @@ task ScatterRegions {
     output {
         Array[File] scatters = glob(outputDirPath + "/scatter-*.bed")
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 2.0]))
+    }
 }
 
 task SampleConfig {
@@ -58,11 +66,15 @@ task SampleConfig {
     String? jsonOutputPath
     String? tsvOutputPath
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        mkdir -p . $(dirname ${jsonOutputPath}) $(dirname ${tsvOutputPath})
-        java -jar ${tool_jar} \
+        mkdir -p . ${"$(dirname " + jsonOutputPath + ")"} ${"$(dirname " + tsvOutputPath + ")"}
+        java -Xmx${mem}G -jar ${tool_jar} \
         -i ${sep="-i " inputFiles} \
         ${"--sample " + sample} \
         ${"--library " + library} \
@@ -77,6 +89,10 @@ task SampleConfig {
         File? tsvOutput = tsvOutputPath
         Object values = if (defined(tsvOutput) && size(tsvOutput) > 0) then read_map(tsvOutput) else { "": "" }
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 2.0]))
+    }
 }
 
 task BaseCounter {
@@ -87,11 +103,15 @@ task BaseCounter {
     String outputDir
     String prefix
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 12.0]))
     command {
         set -e -o pipefail
         ${preCommand}
         mkdir -p ${outputDir}
-        java -jar ${tool_jar} \
+        java -Xmx${mem}G -jar ${tool_jar} \
         -b ${bam} \
         -r ${refFlat} \
         -o ${outputDir} \
@@ -134,4 +154,8 @@ task BaseCounter {
         File transcriptIntronicSense = outputDir + "/" + prefix + ".base.transcript.intronic.sense.counts"
         File transcriptSense = outputDir + "/" + prefix + ".base.transcript.sense.counts"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }

bwa.wdl

@@ -6,15 +6,23 @@ task BwaMem {
     String outputPath
     String? readgroup
 
+    Int? threads
+    Int? memory
+
     command {
         set -e -o pipefail
         mkdir -p $(dirname ${outputPath})
         ${preCommand}
-        bwa mem ${"-R '" + readgroup + "'"} \
+        bwa mem ${"-t " + threads} \
+        ${"-R '" + readgroup + "'"} \
         ${referenceFasta} ${inputR1} ${inputR2} | samtools sort --output-fmt BAM - > ${outputPath}
     }
 
     output {
         File bamFile = outputPath
     }
+    runtime{
+        cpu: if defined(threads) then threads else 1
+        memory: if defined(memory) then memory else 8
+    }
 }

common.wdl

 task objectMd5 {
     Object the_object
+
     command {
         cat ${write_object(the_object)} |  md5sum - | sed -e 's/  -//'
     }
+
     output {
         String md5sum = read_string(stdout())
     }
+
+    runtime {
+        memory: 1
+    }
 }
 
 task mapMd5 {
     Map[String,String] map
+
     command {
-    cat ${write_map(map)} | md5sum - | sed -e 's/  -//'
+        cat ${write_map(map)} | md5sum - | sed -e 's/  -//'
     }
+
     output {
         String md5sum = read_string(stdout())
     }
+
+    runtime {
+        memory: 1
+    }
 }
 
 task stringArrayMd5 {
     Array[String] stringArray
+
     command {
-    set -eu -o pipefail
-    echo ${sep=',' stringArray} | md5sum - | sed -e 's/  -//'
+        set -eu -o pipefail
+        echo ${sep=',' stringArray} | md5sum - | sed -e 's/  -//'
     }
+
     output {
-    String md5sum = read_string(stdout())
+        String md5sum = read_string(stdout())
+    }
+
+    runtime {
+        memory: 1
     }
 }
 
@@ -33,37 +51,68 @@ task concatenateTextFiles {
     Array[File] fileList
     String combinedFilePath
     Boolean? unzip=false
+
     command {
         mkdir -p ${combinedFilePath}
         rm -d ${combinedFilePath}
         ${true='zcat' false= 'cat' unzip} ${sep=' ' fileList} \
         > ${combinedFilePath}
     }
+
     output {
         File combinedFile = combinedFilePath
     }
+
+    runtime {
+        memory: 1
+    }
 }
 
 # inspired by https://gatkforums.broadinstitute.org/wdl/discussion/9616/is-there-a-way-to-flatten-arrays
 task flattenStringArray {
     Array[Array[String]] arrayList
+
     command {
-    for line in $(echo ${sep=', ' arrayList}) ; \
-    do echo $line | tr -d '"[],' ; done
+        for line in $(echo ${sep=', ' arrayList}) ; \
+        do echo $line | tr -d '"[],' ; done
     }
+
     output {
         Array[String] flattenedArray = read_lines(stdout())
     }
+
+    runtime {
+        memory: 1
+    }
 }
 
 task appendToStringArray {
     Array[String] array
     String string
+
     command {
         echo "${sep='\n' array}
         ${string}"
     }
+
     output {
         Array[String] out_array = read_lines(stdout())
     }
+
+    runtime {
+        memory: 1
+    }
+}
+
+task createLink {
+    File inputFile
+    String outputPath
+
+    command {
+        ln -sf ${inputFile} ${outputPath}
+    }
+
+    output {
+        File link = outputPath
+    }
 }
\ No newline at end of file

fastqc.wdl

@@ -62,10 +62,15 @@ task extractAdapters {
     File? knownAdapterFile
     Float? adapterCutoff
     Boolean? outputAsFasta
+
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
     set -e
     mkdir -p ${outputDir}
-    java -jar ${extractAdaptersFastqcJar} \
+    java -Xmx${mem}G -jar ${extractAdaptersFastqcJar} \
     --inputFile ${inputFile} \
     ${"--adapterOutputFile " + adapterOutputFilePath } \
     ${"--contamsOutputFile " + contamsOutputFilePath } \
@@ -82,20 +87,30 @@ task extractAdapters {
         Array[String] adapterList = read_lines(select_first([adapterOutputFilePath]))
         Array[String] contamsList = read_lines(select_first([contamsOutputFilePath]))
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 2.5]))
+    }
 }
 
 task getConfiguration {
     String? preCommand
     String? fastqcDirFile = "fastqcDir.txt"
+
     command {
         set -e -o pipefail
         ${preCommand}
         echo $(dirname $(readlink -f $(which fastqc))) > ${fastqcDirFile}
     }
+
     output {
         String fastqcDir = read_string(fastqcDirFile)
         File adapterList = fastqcDir + "/Configuration/adapter_list.txt"
         File contaminantList = fastqcDir + "/Configuration/contaminant_list.txt"
         File limits = fastqcDir + "/Configuration/limits.txt"
     }
+
+    runtime {
+        memory: 1
+    }
 }
\ No newline at end of file

gatk.wdl

@@ -12,10 +12,14 @@ task BaseRecalibrator {
     File ref_fasta
     File ref_fasta_index
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java -Xms4G -jar ${gatk_jar} \
+        java -Xms${mem}G -jar ${gatk_jar} \
           BaseRecalibrator \
           -R ${ref_fasta} \
           -I ${input_bam} \
@@ -24,9 +28,14 @@ task BaseRecalibrator {
           --known-sites ${sep=" --known-sites " known_indels_sites_VCFs} \
           -L ${sep=" -L " sequence_group_interval}
     }
+
     output {
         File recalibration_report = "${recalibration_report_filename}"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Apply Base Quality Score Recalibration (BQSR) model
@@ -42,10 +51,15 @@ task ApplyBQSR {
     File ref_fasta_index
     Int? compression_level
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xms4G -jar ${gatk_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xms${mem}G -jar ${gatk_jar} \
           ApplyBQSR \
           --create-output-bam-md5 \
           --add-output-sam-program-record \
@@ -57,10 +71,15 @@ task ApplyBQSR {
           --static-quantized-quals 10 --static-quantized-quals 20 --static-quantized-quals 30 \
           -L ${sep=" -L " sequence_group_interval}
     }
+
     output {
         File recalibrated_bam = "${output_bam_path}"
         File recalibrated_bam_checksum = "${output_bam_path}.md5"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Combine multiple recalibration tables from scattered BaseRecalibrator runs
@@ -70,17 +89,26 @@ task GatherBqsrReports {
     Array[File] input_bqsr_reports
     String output_report_filepath
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java -Xms3G -jar ${gatk_jar} \
+        java -Xms${mem}G -jar ${gatk_jar} \
         GatherBQSRReports \
         -I ${sep=' -I ' input_bqsr_reports} \
         -O ${output_report_filepath}
     }
+
     output {
         File output_bqsr_report = "${output_report_filepath}"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Call variants on a single sample with HaplotypeCaller to produce a GVCF
@@ -97,10 +125,15 @@ task HaplotypeCallerGvcf {
     Int? compression_level
     String gatk_jar
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${gatk_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xmx${mem}G -jar ${gatk_jar} \
           HaplotypeCaller \
           -R ${ref_fasta} \
           -O ${gvcf_basename}.vcf.gz \
@@ -109,10 +142,15 @@ task HaplotypeCallerGvcf {
           -contamination ${default=0 contamination} \
           -ERC GVCF
     }
+
     output {
         File output_gvcf = "${gvcf_basename}.vcf.gz"
         File output_gvcf_index = "${gvcf_basename}.vcf.gz.tbi"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 task GenotypeGVCFs {
@@ -133,12 +171,16 @@ task GenotypeGVCFs {
     File dbsnp_vcf_index
 
     Int? compression_level
+    Float? memory
+    Float? memoryMultiplier
 
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
 
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${gatk_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xmx${mem}G -jar ${gatk_jar} \
          GenotypeGVCFs \
          -R ${ref_fasta} \
          -O ${output_basename + ".vcf.gz"} \
@@ -154,6 +196,10 @@ task GenotypeGVCFs {
         File output_vcf = output_basename + ".vcf.gz"
         File output_vcf_index = output_basename + ".vcf.gz.tbi"
     }
+
+    runtime{
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 task CombineGVCFs {
@@ -171,13 +217,17 @@ task CombineGVCFs {
     File ref_dict
 
     Int? compression_level
+    Float? memory
+    Float? memoryMultiplier
 
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
 
         if [ ${length(gvcf_files)} -gt 1 ]; then
-            java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${gatk_jar} \
+            java ${"-Dsamjdk.compression_level=" + compression_level} \
+            -Xmx${mem}G -jar ${gatk_jar} \
              CombineGVCFs \
              -R ${ref_fasta} \
              -O ${output_basename + ".vcf.gz"} \
@@ -193,6 +243,10 @@ task CombineGVCFs {
         File output_gvcf = output_basename + ".vcf.gz"
         File output_gvcf_index = output_basename + ".vcf.gz.tbi"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 task SplitNCigarReads {
@@ -206,11 +260,15 @@ task SplitNCigarReads {
     String gatk_jar
     Array[File]+ intervals
 
+    Float? memory
+    Float? memoryMultiplier
 
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java -Xms4G -jar ${gatk_jar} \
+        java -Xms${mem}G -jar ${gatk_jar} \
+        SplitNCigarReads \
         -I ${input_bam} \
         -R ${ref_fasta} \
         -O ${output_bam} # might have to be -o depending on GATK version \
@@ -221,4 +279,8 @@ task SplitNCigarReads {
         File bam = output_bam
         File bam_index = output_bam + ".bai"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }

htseq.wdl

 task HTSeqCount {
     String? preCommand
     Array[File] alignmentFiles
-    File gffFile
+    File gtfFile
     String outputTable
     String? format
     String? order
     String? stranded
 
+    Int? memory
+
     command {
         set -e -o pipefail
         ${preCommand}
@@ -15,11 +17,15 @@ task HTSeqCount {
         -r ${default="pos" order} \
         -s ${default="no" stranded} \
         ${sep=" " alignmentFiles} \
-        ${gffFile} \
+        ${gtfFile} \
         > ${outputTable}
     }
 
     output {
         File counts = outputTable
     }
+
+    runtime {
+        memory: select_first([memory, 3])
+    }
 }
\ No newline at end of file

mergecounts.wdl

@@ -32,4 +32,8 @@ task MergeCounts {
     output {
         File mergedCounts = outputFile
     }
+
+    runtime {
+        memory: 4 + (2*length(inputFiles))
+    }
 }
\ No newline at end of file

picard.wdl

@@ -4,11 +4,15 @@ task ScatterIntervalList {
     Int scatter_count
     String picard_jar
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
         mkdir scatter_list
-        java -Xmx4G -jar ${picard_jar} \
+        java -Xmx${mem}G -jar ${picard_jar} \
           IntervalListTools \
           SCATTER_COUNT=${scatter_count} \
           SUBDIVISION_MODE=BALANCING_WITHOUT_INTERVAL_SUBDIVISION_WITH_OVERFLOW \
@@ -17,10 +21,15 @@ task ScatterIntervalList {
           INPUT=${interval_list} \
           OUTPUT=scatter_list
     }
+
     output {
         Array[File] out = glob("scatter_list/*/*.interval_list")
         Int interval_count = read_int(stdout())
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Combine multiple recalibrated BAM files from scattered ApplyRecalibration runs
@@ -31,21 +40,31 @@ task GatherBamFiles {
     Int? compression_level
     String picard_jar
 
+    Float? memory
+    Float? memoryMultiplier
+
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${picard_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xmx${mem}G -jar ${picard_jar} \
           GatherBamFiles \
           INPUT=${sep=' INPUT=' input_bams} \
           OUTPUT=${output_bam_path} \
           CREATE_INDEX=true \
           CREATE_MD5_FILE=true
     }
+
     output {
         File output_bam = "${output_bam_path}"
         File output_bam_index = sub(output_bam_path, ".bam$", ".bai")
         File output_bam_md5 = "${output_bam_path}.md5"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Mark duplicate reads to avoid counting non-independent observations
@@ -57,6 +76,9 @@ task MarkDuplicates {
     Int? compression_level
     String picard_jar
 
+    Float? memory
+    Float? memoryMultiplier
+
     # The program default for READ_NAME_REGEX is appropriate in nearly every case.
     # Sometimes we wish to supply "null" in order to turn off optical duplicate detection
     # This can be desirable if you don't mind the estimated library size being wrong and optical duplicate detection is taking >7 days and failing
@@ -65,11 +87,13 @@ task MarkDuplicates {
     # Task is assuming query-sorted input so that the Secondary and Supplementary reads get marked correctly
     # This works because the output of BWA is query-grouped and therefore, so is the output of MergeBamAlignment.
     # While query-grouped isn't actually query-sorted, it's good enough for MarkDuplicates with ASSUME_SORT_ORDER="queryname"
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
         mkdir -p $(dirname ${output_bam_path})
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${picard_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xmx${mem}G -jar ${picard_jar} \
           MarkDuplicates \
           INPUT=${sep=' INPUT=' input_bams} \
           OUTPUT=${output_bam_path} \
@@ -81,11 +105,16 @@ task MarkDuplicates {
           CREATE_INDEX=true \
           ADD_PG_TAG_TO_READS=false
     }
+
     output {
         File output_bam = output_bam_path
         File output_bam_index = sub(output_bam_path, ".bam$", ".bai")
         File duplicate_metrics = metrics_path
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
 
 # Combine multiple VCFs or GVCFs from scattered HaplotypeCaller runs
@@ -97,18 +126,28 @@ task MergeVCFs {
     Int? compression_level
     String picard_jar
 
+    Float? memory
+    Float? memoryMultiplier
+
     # Using MergeVcfs instead of GatherVcfs so we can create indices
     # See https://github.com/broadinstitute/picard/issues/789 for relevant GatherVcfs ticket
+    Int mem = ceil(select_first([memory, 4.0]))
     command {
         set -e -o pipefail
         ${preCommand}
-        java ${"-Dsamjdk.compression_level=" + compression_level} -Xmx4G -jar ${picard_jar} \
+        java ${"-Dsamjdk.compression_level=" + compression_level} \
+        -Xmx${mem}G -jar ${picard_jar} \
           MergeVcfs \
           INPUT=${sep=' INPUT=' input_vcfs} \
           OUTPUT=${output_vcf_path}
     }
+
     output {
         File output_vcf = output_vcf_path
         File output_vcf_index = output_vcf_path + ".tbi"
     }
+
+    runtime {
+        memory: ceil(mem * select_first([memoryMultiplier, 1.5]))
+    }
 }
\ No newline at end of file

samtools.wdl

@@ -21,7 +21,12 @@ task Merge {
     command {
         set -e -o pipefail
         ${preCommand}
-        samtools merge ${outputBamPath} ${sep=' ' bamFiles}
+        if [ ${length(bamFiles)} -gt 1 ]
+          then
+            samtools merge ${outputBamPath} ${sep=' ' bamFiles}
+          else
+            ln -sf ${bamFiles} ${outputBamPath}
+        fi
     }
 
     output {

star.wdl

@@ -2,7 +2,7 @@ task Star {
     String? preCommand
 
     Array[File] inputR1
-    Array[File]? inputR2
+    Array[File?] inputR2
     String genomeDir
     String outFileNamePrefix
 
@@ -13,9 +13,14 @@ task Star {
     String? twopassMode
     Array[String]? outSAMattrRGline
 
+    Int? memory
+
     #TODO needs to be extended for all possible output extensions
     Map[String, String] samOutputNames = {"BAM SortedByCoordinate": "sortedByCoord.out.bam"}
 
+    # converts String? to String for use as key (for the Map above) in output
+    String key = select_first([outSAMtype, "BAM SortedByCoordinate"])
+
     command {
         set -e -o pipefail
         mkdir -p ${sub(outFileNamePrefix, basename(outFileNamePrefix) + "$", "")}
@@ -33,10 +38,11 @@ task Star {
     }
 
     output {
-        File bamFile = outFileNamePrefix + "Aligned." +  samOutputNames["${outSAMtype}"]
+        File bamFile = outFileNamePrefix + "Aligned." +  samOutputNames[key]
     }
 
     runtime {
-        threads: runThreadN
+        cpu: select_first([runThreadN, 1])
+        memory: select_first([memory, 10])
     }
 }
\ No newline at end of file

stringtie.wdl

 task Stringtie {
     String? preCommand
     File alignedReads
-    File? referenceGFF
+    File? referenceGtf
     Int? threads
     String assembledTranscriptsFile
     Boolean? firstStranded
@@ -13,7 +13,7 @@ task Stringtie {
         ${preCommand}
         stringtie \
         ${"-p " + threads} \
-        ${"-G " + referenceGFF} \
+        ${"-G " + referenceGtf} \
         ${true="--rf" false="" firstStranded} \
         ${true="fr" false="" secondStranded} \
         -o ${assembledTranscriptsFile} \
@@ -28,6 +28,6 @@ task Stringtie {
     }
 
     runtime {
-        threads: threads
+        cpu: select_first([threads, 1])
     }
 }
\ No newline at end of file