From 50c5d957408dbf8a6f1d6aa79c0a3b05ffdde664 Mon Sep 17 00:00:00 2001
From: Ruben Vorderman <r.h.p.vorderman@lumc.nl>
Date: Tue, 7 Jul 2020 12:41:49 +0200
Subject: [PATCH] Add samtools controls to hisat2
---
CHANGELOG.md | 3 +++
hisat2.wdl | 21 +++++++++++++++------
2 files changed, 18 insertions(+), 6 deletions(-)
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 4acadc5..edfffb5 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -11,6 +11,9 @@ that users understand how the changes affect the new version.
version 4.0.0-develop
---------------------------
++ Hisat2 task has added controls for samtools.
++ Alignment tasks no longer produce BAM indexes as these are not needed
+ by the markduplicates step.
+ Picard SortSam added as a task.
+ Md5 files are no longer created by default on Picard tasks that generate
BAM files.
diff --git a/hisat2.wdl b/hisat2.wdl
index 5937f86..77c370f 100644
--- a/hisat2.wdl
+++ b/hisat2.wdl
@@ -34,7 +34,10 @@ task Hisat2 {
String summaryFilePath = basename(outputBam, ".bam") + ".summary.txt"
Int threads = 4
- String memory = "~{threads + 5 + ceil(size(indexFiles, "G"))}G"
+ Int sortThreads = 1
+ Int sortMemoryPerThreadGb = 2
+ Int compressionLevel = 1
+ Int memoryGb = 1 + threads + ceil(size(indexFiles, "G") * 1.2) + sortMemoryPerThreadGb * sortThreads
Int timeMinutes = 1 + ceil(size([inputR1, inputR2], "G") * 180 / threads)
# quay.io/biocontainers/mulled-v2-a97e90b3b802d1da3d6958e0867610c718cb5eb1
# is a combination of hisat2 and samtools
@@ -59,18 +62,21 @@ task Hisat2 {
~{true="--dta" false="" downstreamTranscriptomeAssembly} \
--new-summary \
--summary-file ~{summaryFilePath} \
- | samtools sort > ~{outputBam}
- samtools index ~{outputBam} ~{bamIndexPath}
+ | samtools sort \
+ ~{"-@ " + sortThreads} \
+ -m ~{sortMemoryPerThreadGb}G \
+ -l ~{compressionLevel} \
+ - \
+ -o ~{outputBam}
}
output {
File bamFile = outputBam
- File bamIndex = bamIndexPath
File summaryFile = summaryFilePath
}
runtime {
- memory: memory
+ memory: "~{memoryGb}G"
cpu: threads + 1
time_minutes: timeMinutes
docker: dockerImage
@@ -88,9 +94,12 @@ task Hisat2 {
downstreamTranscriptomeAssembly: {description: "Equivalent to hisat2's `--dta` flag.", category: "advanced"}
summaryFilePath: {description: "Where the summary file should be written.", category: "advanced"}
threads: {description: "The number of threads to use.", category: "advanced"}
- memory: {description: "The amount of memory this job will use.", category: "advanced"}
+ memoryGb: {description: "The amount of memory this job will use in gigabytes.", category: "advanced"}
timeMinutes: {description: "The maximum amount of time the job will run in minutes.", category: "advanced"}
dockerImage: {description: "The docker image used for this task. Changing this may result in errors which the developers may choose not to address.",
category: "advanced"}
+ sortThreads: {description: "The number of threads to use for sorting.", category: "advanced"}
+ sortMemoryPerThreadGb: {description: "The amount of memory for each sorting thread in gigabytes.", category: "advanced"}
+ compressionLevel: {description: "The compression level of the output BAM.", category: "advanced"}
}
}
\ No newline at end of file
--
GitLab