From 1fe7d8b3ffc800bc16f79f193c5a7e423c993033 Mon Sep 17 00:00:00 2001
From: "Gervaise H. Henry" <gervaise.henry@utsouthwestern.edu>
Date: Wed, 20 Jan 2021 11:56:21 -0600
Subject: [PATCH] Use 1 less process for samtools threading and limit memto 75%
of available
---
CHANGELOG.md | 5 +++--
workflow/rna-seq.nf | 10 ++++++----
2 files changed, 9 insertions(+), 6 deletions(-)
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 761a706..b4dfcdf 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -2,7 +2,7 @@
**User Facing**
**Background**
-* Add memory limit per thread for samtools sort (#108)
+* Add memory limit (75%) per thread for samtools sort (#108)
* Remove parsing restrictions for submitted stranded/spike/species (#105, #106)
* Pass unidentified ends instead of overwriting it as unknown
* Move fastqc process before trim to catch fastq errors (#107)
@@ -13,7 +13,8 @@
* Handle blank submitted endness better
* Don't use file.csv from inputBag to parse manual endness, use counted from getData
* Detect malformed fastq's (#107)
-* Restrict sampled alignment process to use >32GB nodes on BioHPC
+* Restrict sampled alignment process to use >32GB nodes on BioHPC (#108)
+* Use nproc**-1** for alignment processes (#108)
*Known Bugs*
* Override params (inputBag, fastq, species) aren't checked for integrity
diff --git a/workflow/rna-seq.nf b/workflow/rna-seq.nf
index 5c68c00..e596205 100644
--- a/workflow/rna-seq.nf
+++ b/workflow/rna-seq.nf
@@ -873,9 +873,10 @@ process alignSampleData {
# sort the bam file using Samtools
echo -e "LOG: sorting the bam file" >> ${repRID}.${ref}.alignSampleData.log
+ proc=\$(expr `nproc` - 1)
mem=\$(vmstat -s -S K | grep 'total memory' | grep -o '[0-9]*')
- mem=\$(expr \${mem} / `nproc` \\* 85 / 100)
- samtools sort -@ `nproc` -m \${mem}K -O BAM -o ${ref}.sampled.sorted.bam ${ref}.sampled.bam
+ mem=\$(expr \${mem} / \${proc} \\* 85 / 100)
+ samtools sort -@ \${proc} -m \${mem}K -O BAM -o ${ref}.sampled.sorted.bam ${ref}.sampled.bam
# index the sorted bam using Samtools
echo -e "LOG: indexing sorted bam file" >> ${repRID}.${ref}.alignSampleData.log
@@ -1576,9 +1577,10 @@ process alignData {
# sort the bam file using Samtools
echo -e "LOG: sorting the bam file" >> ${repRID}.align.log
+ proc=\$(expr `nproc` - 1)
mem=\$(vmstat -s -S K | grep 'total memory' | grep -o '[0-9]*')
- mem=\$(expr \${mem} / `nproc` \\* 85 / 100)
- samtools sort -@ `nproc` -m \${mem}K -O BAM -o ${repRID}.sorted.bam ${repRID}.bam
+ mem=\$(expr \${mem} / \${proc} \\* 75 / 100)
+ samtools sort -@ \${proc} -m \${mem}K -O BAM -o ${repRID}.sorted.bam ${repRID}.bam
# index the sorted bam using Samtools
echo -e "LOG: indexing sorted bam file" >> ${repRID}.align.log
--
GitLab