Split java opts from mem resources in config

config/config.yaml

@@ -671,11 +671,21 @@ params:
     # system-provided tmp dir is too small (which can happen on clusters).
     # Note that the Java memory options, such as `-Xmx10g` to increase the available memory within
     # the Java virtual machine are provided via the Snakemake memory management directly,
-    # for instance, via the memory specified in slurm jobs. Change the memory there as needed.
+    # and hence cannot be specified here. Instead, use the below `*-mem-mb` options,
+    # or, if you are running grenepipe via slurm, use the slurm job configuration.
     # The last option, SortVcf-java-opts, is used by bcftools when using contig-group-size > 0.
     MarkDuplicates-java-opts: ""
     CollectMultipleMetrics-java-opts: ""
     SortVcf-java-opts: ""
+    MergeVcfs-java-opts: ""
+
+    # Memory for the Java virtual machine for the picard programs.
+    # Unfortunately, Java does not automatically use the available memory, and instead needs
+    # to be told that it is allowed to do that. Specify the memory here as needed, in MB.
+    MarkDuplicates-mem-mb: 5000
+    CollectMultipleMetrics-mem-mb: 1024
+    SortVcf-mem-mb: 1024
+    MergeVcfs-mem-mb: 1024
 
   # ----------------------------------------------------------------------
   #     dedup
@@ -770,10 +780,21 @@ params:
     GenotypeGVCFs-extra: ""
 
     # For some specific error cases, it might be necessary to adjust java settings for the tools.
+    # Note that the Java memory options, such as `-Xmx10g` to increase the available memory within
+    # the Java virtual machine are provided via the Snakemake memory management directly,
+    # and hence cannot be specified here. Instead, use the below `*-mem-mb` options,
+    # or, if you are running grenepipe via slurm, use the slurm job configuration.
     HaplotypeCaller-java-opts: ""
     CombineGVCFs-java-opts: ""
     GenotypeGVCFs-java-opts: ""
 
+    # Memory for the Java virtual machine for the GATK programs.
+    # Unfortunately, Java does not automatically use the available memory, and instead needs
+    # to be told that it is allowed to do that. Specify the memory here as needed, in MB.
+    HaplotypeCaller-mem-mb: 1024
+    CombineGVCFs-mem-mb: 1024
+    GenotypeGVCFs-mem-mb: 1024
+
     # Number of threads to use for the HaplotypeCaller. We recommend to keep this at 2,
     # as GATK does not seem to do a great job of parallelizing anyway.
     HaplotypeCaller-threads: 2
@@ -797,6 +818,7 @@ params:
     # We also offer extra settings that are used for both.
     extra: ""
     java-opts: ""
+    mem-mb: 1024
 
   # ----------------------------------------------------------------------
   #     GATK VariantRecalibrator + ApplyVQSR
@@ -878,14 +900,16 @@ params:
     # We here set the gaussians to a very low value, simply for our test case to work, see
     # https://gatk.broadinstitute.org/hc/en-us/community/posts/4408833794587-Insufficient-variance-error-for-VariantRecalibrator
     # for details. We recommend to change this according to your actual needs.
-    extra-variantrecalibrator-SNP: "--max-gaussians 1"
-    extra-variantrecalibrator-INDEL: "--max-gaussians 1"
-    java-variantrecalibrator: ""
+    variantrecalibrator-extra-SNP: "--max-gaussians 1"
+    variantrecalibrator-extra-INDEL: "--max-gaussians 1"
+    variantrecalibrator-java-opts: ""
+    variantrecalibrator-mem-mb: 1024
 
     # Extra command line params, and optional Java runtime options to provide to GATK ApplyVQSR
-    extra-applyvqsr-SNP: "--truth-sensitivity-filter-level 99.0"
-    extra-applyvqsr-INDEL: "--truth-sensitivity-filter-level 99.0"
-    java-applyvqsr: ""
+    applyvqsr-extra-SNP: "--truth-sensitivity-filter-level 99.0"
+    applyvqsr-extra-INDEL: "--truth-sensitivity-filter-level 99.0"
+    applyvqsr-java-opts: ""
+    applyvqsr-mem-mb: 1024
 
   # ----------------------------------------------------------------------
   #     bcftools filter

workflow/rules/calling-bcftools.smk

@@ -76,7 +76,7 @@ if config["settings"].get("contig-group-size"):
                 if platform.system() == "Darwin"
                 else ""
             ),
-            java_opts=config["params"]["picard"]["SortVcf-java-opts"],
+            java_opts=config["params"]["picard"]["SortVcf-java-opts"] + " -Xmx" + config["params"]["picard"].get("SortVcf-mem-mb", 1024) + "m",
         log:
             "logs/calling/picard/sort-genotyped.log",
         benchmark:
@@ -87,8 +87,8 @@ if config["settings"].get("contig-group-size"):
             # Weird new picard syntax...
             "picard SortVcf "
             "{params.java_opts} "
-            "INPUT={input.vcf} "
-            "OUTPUT={output.vcf} "
-            "SEQUENCE_DICTIONARY={input.refdict} "
+            "--INPUT {input.vcf} "
+            "--OUTPUT {output.vcf} "
+            "--SEQUENCE_DICTIONARY {input.refdict} "
             "{params.extra} "
             "> {log} 2>&1"

workflow/rules/calling-haplotypecaller.smk

@@ -77,13 +77,15 @@ rule call_variants:
         # Contigs are used as long as no restrict-regions are given in the config file.
         extra=get_gatk_call_variants_params,
         java_opts=config["params"]["gatk"]["HaplotypeCaller-java-opts"],
+    resources:
+        mem_mb=config["params"]["gatk"].get("HaplotypeCaller-mem-mb", 1024),
     group:
         "call_variants"
     conda:
         # Need to specify, yet again...
         "../envs/gatk.yaml"
     wrapper:
-        "0.51.3/bio/gatk/haplotypecaller"
+        "v5.7.0/bio/gatk/haplotypecaller"
 
 
 # Deactivated the below, as this was causing trouble. Got the warning
@@ -155,6 +157,8 @@ rule combine_calls:
             else ""
         ),
         java_opts=config["params"]["gatk"]["CombineGVCFs-java-opts"],
+    resources:
+        mem_mb=config["params"]["gatk"].get("CombineGVCFs-mem-mb", 1024),
     log:
         "logs/calling/gatk-combine-gvcfs/{contig}.log",
     benchmark:
@@ -164,7 +168,7 @@ rule combine_calls:
     conda:
         "../envs/gatk.yaml"
     wrapper:
-        "0.51.3/bio/gatk/combinegvcfs"
+        "v5.7.0/bio/gatk/combinegvcfs"
 
 
 rule genotype_variants:
@@ -194,6 +198,8 @@ rule genotype_variants:
             else ""
         ),
         java_opts=config["params"]["gatk"]["GenotypeGVCFs-java-opts"],
+    resources:
+        mem_mb=config["params"]["gatk"].get("GenotypeGVCFs-mem-mb", 1024),
     log:
         "logs/calling/gatk-genotype-gvcfs/{contig}.log",
     benchmark:
@@ -203,7 +209,7 @@ rule genotype_variants:
     conda:
         "../envs/gatk.yaml"
     wrapper:
-        "0.51.3/bio/gatk/genotypegvcfs"
+        "v5.7.0/bio/gatk/genotypegvcfs"
 
 
 # =================================================================================================
@@ -239,14 +245,17 @@ rule merge_variants:
         done=touch("calling/genotyped-all.vcf.gz.done"),
     params:
         # See duplicates-picard.smk for the reason whe need this on MacOS.
+        java_opts=config["params"]["picard"]["MergeVcfs-java-opts"],
         extra=(
             " --USE_JDK_DEFLATER true --USE_JDK_INFLATER true" if platform.system() == "Darwin" else ""
         ),
+    resources:
+        mem_mb=config["params"]["picard"].get("MergeVcfs-mem-mb", 1024),
     log:
         "logs/calling/picard-merge-genotyped.log",
     benchmark:
         "benchmarks/calling/genotyped/picard/merge-genotyped.log"
     conda:
         "../envs/picard.yaml"
     wrapper:
-        "0.51.3/bio/picard/mergevcfs"
+        "v5.7.0/bio/picard/mergevcfs"

workflow/rules/duplicates-picard.smk

@@ -36,8 +36,8 @@ rule mark_duplicates:
         extra=config["params"]["picard"]["MarkDuplicates"]
         + (" --USE_JDK_DEFLATER true --USE_JDK_INFLATER true" if platform.system() == "Darwin" else ""),
         java_opts=config["params"]["picard"]["MarkDuplicates-java-opts"]
-    # resources:
-    #     mem_mb=1024,
+    resources:
+        mem_mb=config["params"]["picard"].get("MarkDuplicates-mem-mb", 5000),
     group:
         "mapping_extra"
     conda:

workflow/rules/filtering-gatk-variantfiltration.smk

@@ -29,6 +29,8 @@ rule gatk_hard_filter_calls:
         filters=get_filter,
         extra=config["params"]["gatk-variantfiltration"]["extra"],
         java_opts=config["params"]["gatk-variantfiltration"]["java-opts"],
+    resources:
+        mem_mb=config["params"]["gatk-variantfiltration"].get("mem-mb", 1024),
     log:
         "logs/calling/gatk-variantfiltration/{vartype}.log",
     benchmark:
@@ -38,4 +40,4 @@ rule gatk_hard_filter_calls:
     conda:
         "../envs/gatk.yaml"
     wrapper:
-        "0.85.0/bio/gatk/variantfiltration"
+        "v5.7.0/bio/gatk/variantfiltration"

workflow/rules/filtering-gatk-vqsr.smk

@@ -57,7 +57,7 @@ def get_variant_recalibrator_extra(wildcards):
     # config file dict retrieval. We additionally set the Rscript file, so that we get some plots.
     vartype = wildcards.vartype
     return (
-        config["params"]["gatk-vqsr"]["extra-variantrecalibrator-" + vartype]
+        config["params"]["gatk-vqsr"]["variantrecalibrator-extra-" + vartype]
         + " --rscript-file calling/filtered/all."
         + vartype
         + ".vqsr-recal.plots.R"
@@ -66,7 +66,7 @@ def get_variant_recalibrator_extra(wildcards):
 
 def get_apply_vqsr_extra(wildcards):
     # Same as above, need a function here for wildcard replacement
-    return config["params"]["gatk-vqsr"]["extra-applyvqsr-" + wildcards.vartype]
+    return config["params"]["gatk-vqsr"]["applyvqsr-extra-" + wildcards.vartype]
 
 
 # Use the GATK VQSR machine learning based recalibration of quality scores instead of hard filtering.
@@ -100,7 +100,7 @@ rule gatk_variant_recalibrator:
         annotation=config["params"]["gatk-vqsr"]["annotation"],
         # Extras
         extra=get_variant_recalibrator_extra,
-        java_opts=config["params"]["gatk-vqsr"]["java-variantrecalibrator"],
+        java_opts=config["params"]["gatk-vqsr"]["variantrecalibrator-java-opts"] + " -Xmx" + config["params"]["gatk-vqsr"].get("variantrecalibrator-mem-mb", 1024) + "m",
     log:
         "logs/calling/gatk-variantrecalibrator/{vartype}.log",
     benchmark:
@@ -147,7 +147,7 @@ rule gatk_apply_vqsr:
         # set mode, must be either SNP, INDEL or BOTH
         mode="{vartype}",
         extra=get_apply_vqsr_extra,
-        java_opts=config["params"]["gatk-vqsr"]["java-applyvqsr"],
+        java_opts=config["params"]["gatk-vqsr"]["applyvqsr-java-opts"] + " -Xmx" + config["params"]["gatk-vqsr"].get("applyvqsr-mem-mb", 1024) + "m",
     # resources:
     # mem_mb=50
     conda:

workflow/rules/qc-bam.smk

@@ -230,6 +230,8 @@ rule picard_collectmultiplemetrics:
         + " "
         + config["params"]["picard"]["CollectMultipleMetrics-extra"]
         + (" --USE_JDK_DEFLATER true --USE_JDK_INFLATER true" if platform.system() == "Darwin" else ""),
+    resources:
+        mem_mb=config["params"]["picard"].get("CollectMultipleMetrics-mem-mb", 1024),
     conda:
         "../envs/picard.yaml"
     # script:

workflow/schemas/config.schema.yaml

@@ -321,17 +321,17 @@ properties:
             type: [array, object]
           annotation:
             type: [array, object]
-          extra-variantrecalibrator-SNP:
+          variantrecalibrator-extra-SNP:
             type: string
-          extra-variantrecalibrator-INDEL:
+          variantrecalibrator-extra-INDEL:
             type: string
-          java-variantrecalibrator:
+          variantrecalibrator-java-opts:
             type: string
-          extra-applyvqsr-SNP:
+          applyvqsr-extra-SNP:
             type: string
-          extra-applyvqsr-INDEL:
+          applyvqsr-extra-INDEL:
             type: string
-          java-applyvqsr:
+          applyvqsr-java-opts:
             type: string
       bcftools-filter:
         type: object