Merge pull request #549 from broadinstitute/ct-increase-trinity-mem

raise trinity rule memory 4g->10g, JVMmemory 5g
broadinstitute · Dec 14, 2016 · 2e9bd5f · 2e9bd5f
2 parents 7319dda + bce25ae
commit 2e9bd5f
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 4 deletions.
diff --git a/pipes/rules/assembly.rules b/pipes/rules/assembly.rules
@@ -39,7 +39,7 @@ rule assemble_trinity:
     input:  config["data_dir"]+'/'+config["subdirs"]["per_sample"]+'/{sample}.taxfilt.bam'
     output: config["tmp_dir"] +'/'+config["subdirs"]["assembly"]+'/{sample}.assembly1-trinity.fasta'
     resources: 
-            mem=4,
+            mem=7,
             cores=int(config.get("number_of_threads", 1))
     params: LSF=config.get('LSF_queues', {}).get('short', '-W 4:00'),
             UGER=config.get('UGER_queues', {}).get('short', '-q short'),
@@ -52,7 +52,7 @@ rule assemble_trinity:
             makedirs(expand("{dir}/{subdir}",
                 dir=[config["data_dir"],config["tmp_dir"]],
                 subdir=config["subdirs"]["assembly"]))
-            shell("{config[bin_dir]}/assembly.py assemble_trinity {input} {params.clipDb} {output} --n_reads={params.n_reads} --outReads {params.subsamp_bam} --threads {params.numThreads}")
+            shell("{config[bin_dir]}/assembly.py assemble_trinity {input} {params.clipDb} {output} --n_reads={params.n_reads} --outReads {params.subsamp_bam} --threads {params.numThreads} --JVMmemory 5g")
 
 rule orient_and_impute:
     ''' This step cleans up the Trinity assembly with a known reference genome.

diff --git a/pipes/rules/interhost.rules b/pipes/rules/interhost.rules
@@ -91,6 +91,16 @@ rule ref_guided_diversity:
             shell("cat " + " ".join(params.inFastas) + " > {output[0]}")
             merge_vcfs(params.inVcfs, params.refGenome, output[1])
 
+def multi_align_mafft_memory(_, input):
+    sample_count = len(list(read_samples_file(config["samples_assembly"])))
+    if sample_count in range(0,50):
+        mem=8
+    elif sample_count in range(50,100):
+        mem=16
+    elif sample_count >= 100:
+        mem=32
+    return mem
+
 rule multi_align_mafft:
     input:
         expand("{data_dir}/{subdir}/{sample}.fasta",
@@ -105,10 +115,10 @@ rule multi_align_mafft:
                 subdir=config["subdirs"]["multialign_ref"],
                 chrom=range(1, len(config["accessions_for_ref_genome_build"])+1))        
     resources: 
-            mem=8,
+            mem=multi_align_mafft_memory,
             cores=int(config.get("number_of_threads", 1))
     params: LSF=config.get('LSF_queues', {}).get('short', '-W 4:00'),
-            UGER=config.get('UGER_queues', {}).get('short', '-q short'),
+            UGER=config.get('UGER_queues', {}).get('long', '-q long'),
             logid="all",
             refGenome=os.path.join(config["ref_genome_dir"],"reference"+".fasta"),
             samples=list(read_samples_file(config["samples_assembly"])),