pulled from chip pipeline code from ihec-assay-standards/commit/40aa3…

…1f8b715d5215d162770863a3260706bdb0c
IHEC · Apr 1, 2019 · f10e902 · f10e902
1 parent ac161ea
commit f10e902
Show file tree

Hide file tree

Showing 24 changed files with 1,233 additions and 2 deletions.
diff --git a/ChIP-Seq_pipeline_proposal.pdf b/ChIP-Seq_pipeline_proposal.pdf
diff --git a/README.md b/README.md
diff --git a/encode-wrapper/backend.conf b/encode-wrapper/backend.conf
@@ -0,0 +1,250 @@
+include required(classpath("application"))
+
+backend {
+  default = "Local"
+  providers {
+
+    slurm_singularity {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        script-epilogue = "sleep 30"
+        concurrent-job-limit = 50
+        runtime-attributes = """
+          Int cpu = 1
+          Int? gpu
+          Int? time
+          Int? memory_mb
+          String? slurm_partition
+          String? slurm_account
+          String? slurm_extra_param
+          String singularity_container
+          String? singularity_bindpath
+        """
+        submit = """
+          ls ${singularity_container} $(echo ${singularity_bindpath} | tr , ' ') 1>/dev/null && (sbatch \
+          --export=ALL \
+          -J ${job_name} \
+          -D ${cwd} \
+          -o ${out} \
+          -e ${err} \
+          ${"-t " + time*60} \
+          -n 1 \
+          --ntasks-per-node=1 \
+          ${"--cpus-per-task=" + cpu} \
+          ${"--mem=" + memory_mb} \
+          ${"-p " + slurm_partition} \
+          ${"--account " + slurm_account} \
+          ${"--gres gpu:" + gpu} \
+          ${slurm_extra_param} \
+          --wrap "chmod u+x ${script} && SINGULARITY_BINDPATH=$(echo ${cwd} | sed 's/cromwell-executions/\n/g' | head -n1),${singularity_bindpath} singularity exec --home ${cwd} ${if defined(gpu) then '--nv' else ''} ${singularity_container} ${script}")
+        """
+        kill = "scancel ${job_id}"
+        check-alive = "squeue -j ${job_id}"
+        job-id-regex = "Submitted batch job (\\d+).*"
+      }
+    }
+
+    sge_singularity {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        script-epilogue = "sleep 30 && sync"
+        concurrent-job-limit = 50
+        runtime-attributes = """
+          String sge_pe = "shm"
+          Int cpu = 1
+          Int? gpu
+          Int? time
+          Int? memory_mb
+          String? sge_queue
+          String? sge_extra_param
+          String singularity_container
+          String? singularity_bindpath
+        """
+        submit = """
+          ls ${singularity_container} $(echo ${singularity_bindpath} | tr , ' ') 1>/dev/null && (echo "chmod u+x ${script} && SINGULARITY_BINDPATH=$(echo ${cwd} | sed 's/cromwell-executions/\n/g' | head -n1),${singularity_bindpath} singularity exec --home ${cwd} ${if defined(gpu) then '--nv' else ''} ${singularity_container} ${script}" | qsub \
+          -S /bin/sh \
+          -terse \
+          -b n \
+          -N ${job_name} \
+          -wd ${cwd} \
+          -o ${out} \
+          -e ${err} \
+          ${if cpu>1 then "-pe " + sge_pe + " " + cpu else " "} \
+          ${"-l h_vmem=" + memory_mb/cpu + "m"} \
+          ${"-l s_vmem=" + memory_mb/cpu + "m"} \
+          ${"-l h_rt=" + time*3600} \
+          ${"-l s_rt=" + time*3600} \
+          ${"-q " + sge_queue} \
+          ${"-l gpu=" + gpu} \
+          ${sge_extra_param} \
+          -V)
+        """
+        kill = "qdel ${job_id}"
+        check-alive = "qstat -j ${job_id}"
+        job-id-regex = "(\\d+)"
+      }
+    }
+
+    singularity {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        script-epilogue = "sleep 5 && sync"
+        concurrent-job-limit = 10
+        runtime-attributes = """
+          Int? gpu
+          String singularity_container
+          String? singularity_bindpath
+        """
+        submit = """
+          ls ${singularity_container} $(echo ${singularity_bindpath} | tr , ' ') 1>/dev/null && (chmod u+x ${script} && SINGULARITY_BINDPATH=$(echo ${cwd} | sed 's/cromwell-executions/\n/g' | head -n1),${singularity_bindpath} singularity exec --home ${cwd} ${if defined(gpu) then '--nv' else ''} ${singularity_container} ${script} & echo $! && disown)
+        """
+        job-id-regex = "(\\d+)"
+        check-alive = "ps -ef | grep -v grep | grep ${job_id}" 
+        kill = "kill -9 ${job_id}"
+      }
+    }
+
+    Local {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        concurrent-job-limit = 10
+      }
+    }
+
+    sge {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        script-epilogue = "sleep 30 && sync"
+        concurrent-job-limit = 50
+        runtime-attributes = """
+        String sge_pe = "shm"
+        Int cpu = 1
+        Int? gpu
+        Int? time
+        Int? memory_mb
+        String? sge_queue
+        String? sge_extra_param
+        """
+        submit = """
+        qsub \
+        -S /bin/sh \
+        -terse \
+        -b n \
+        -N ${job_name} \
+        -wd ${cwd} \
+        -o ${out} \
+        -e ${err} \
+        ${if cpu>1 then "-pe " + sge_pe + " " + cpu else " "} \
+        ${"-l h_vmem=" + memory_mb/cpu + "m"} \
+        ${"-l s_vmem=" + memory_mb/cpu + "m"} \
+        ${"-l h_rt=" + time*3600} \
+        ${"-l s_rt=" + time*3600} \
+        ${"-q " + sge_queue} \
+        ${"-l gpu=" + gpu} \
+        ${sge_extra_param} \
+        -V \
+        ${script}
+        """
+        kill = "qdel ${job_id}"
+        check-alive = "qstat -j ${job_id}"
+        job-id-regex = "(\\d+)"
+      }
+    }
+
+    slurm {
+      actor-factory = "cromwell.backend.impl.sfs.config.ConfigBackendLifecycleActorFactory"
+      config {
+        script-epilogue = "sleep 30"
+        concurrent-job-limit = 50        
+        runtime-attributes = """
+        Int cpu = 1
+        Int? gpu
+        Int? time
+        Int? memory_mb
+        String? slurm_partition
+        String? slurm_account
+        String? slurm_extra_param
+        """
+        submit = """
+        sbatch \
+        --export=ALL \
+        -J ${job_name} \
+        -D ${cwd} \
+        -o ${out} \
+        -e ${err} \
+        ${"-t " + time*60} \
+        -n 1 \
+        --ntasks-per-node=1 \
+        ${"--cpus-per-task=" + cpu} \
+        ${"--mem=" + memory_mb} \
+        ${"-p " + slurm_partition} \
+        ${"--account " + slurm_account} \
+        ${"--gres gpu:" + gpu} \
+        ${slurm_extra_param} \
+        --wrap "/bin/bash ${script}"
+        """
+        kill = "scancel ${job_id}"
+        check-alive = "squeue -j ${job_id}"
+        job-id-regex = "Submitted batch job (\\d+).*"
+      }
+    }
+
+    google {
+      actor-factory = "cromwell.backend.impl.jes.JesBackendLifecycleActorFactory"
+      config {
+        # Google project
+        project = "your-project-name"
+
+        # Base bucket for workflow executions
+        root = "gs://your-bucket-name"
+
+        concurrent-job-limit = 1000
+        genomics-api-queries-per-100-seconds = 1000
+        maximum-polling-interval = 600
+
+        genomics {
+          auth = "application-default"
+          compute-service-account = "default"
+          endpoint-url = "https://genomics.googleapis.com/"
+          restrict-metadata-access = false
+        }
+
+        filesystems {
+          gcs {
+            auth = "application-default"
+          }
+        }
+      }
+    }
+  }
+}
+
+services {
+  LoadController {
+    class = "cromwell.services.loadcontroller.impl.LoadControllerServiceActor"
+    config {      
+      # disable it (for login nodes on Stanford SCG, Sherlock)
+      control-frequency = 21474834 seconds
+    }
+  }
+}
+
+system {
+  abort-jobs-on-terminate = true
+  graceful-server-shutdown = true
+}
+
+call-caching {
+  enabled = false
+  invalidate-bad-cache-results = true
+}
+
+google {
+  application-name = "cromwell"
+  auths = [
+    {
+      name = "application-default"
+      scheme = "application_default"
+    }
+  ]
+}
diff --git a/encode-wrapper/cemt0007_h3k27me3_mnt_ext_0.json b/encode-wrapper/cemt0007_h3k27me3_mnt_ext_0.json
@@ -0,0 +1,24 @@
+{
+    "chip.choose_ctl.always_use_pooled_ctl": true, 
+    "chip.ctl_fastqs": [
+        [
+            [
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.1.fastq.gz", 
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.2.fastq.gz"
+            ]
+        ]
+    ], 
+    "chip.fastqs": [
+        [
+            [
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26688-GGCTAC.134224.D2B0LACXX.2.1.fastq.gz", 
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26688-GGCTAC.134224.D2B0LACXX.2.2.fastq.gz"
+            ]
+        ]
+    ], 
+    "chip.genome_tsv": "/mnt/ext_0/test_genome_database/hg38_local.tsv", 
+    "chip.paired_end": true, 
+    "chip.pipeline_type": "histone", 
+    "chip.qc_report.desc": "CEMT0007_MCF10A H3K27me3", 
+    "chip.qc_report.name": "CEMT0007_MCF10A-H3K27me3"
+}
diff --git a/encode-wrapper/cemt0007_h3k27me3_template.json b/encode-wrapper/cemt0007_h3k27me3_template.json
@@ -0,0 +1,16 @@
+{
+    "chip.pipeline_type" : "histone",
+    "chip.genome_tsv" : "{0}/test_genome_database/hg38_local.tsv",
+    "chip.fastqs" : [
+        [["{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26688-GGCTAC.134224.D2B0LACXX.2.1.fastq.gz",
+          "{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26688-GGCTAC.134224.D2B0LACXX.2.2.fastq.gz"]]
+    ],
+    "chip.ctl_fastqs" : [
+        [["{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.1.fastq.gz",
+          "{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.2.fastq.gz"]]
+    ],
+    "chip.paired_end" : true,
+    "chip.choose_ctl.always_use_pooled_ctl" : true,
+    "chip.qc_report.name" : "CEMT0007_MCF10A-H3K27me3",
+    "chip.qc_report.desc" : "CEMT0007_MCF10A H3K27me3"
+}
diff --git a/encode-wrapper/cemt0007_h3k4me3_mnt_ext_0.json b/encode-wrapper/cemt0007_h3k4me3_mnt_ext_0.json
@@ -0,0 +1,24 @@
+{
+    "chip.choose_ctl.always_use_pooled_ctl": true, 
+    "chip.ctl_fastqs": [
+        [
+            [
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.1.fastq.gz", 
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.2.fastq.gz"
+            ]
+        ]
+    ], 
+    "chip.fastqs": [
+        [
+            [
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A28471-ATCACG.134224.D2B0LACXX.2.1.fastq.gz", 
+                "/mnt/ext_1/test_data/ChIP-Seq.IX1239-A28471-ATCACG.134224.D2B0LACXX.2.2.fastq.gz"
+            ]
+        ]
+    ], 
+    "chip.genome_tsv": "/mnt/ext_0/test_genome_database/hg38_local.tsv", 
+    "chip.paired_end": true, 
+    "chip.pipeline_type": "histone", 
+    "chip.qc_report.desc": "CEMT0007_MCF10A H3K4me3", 
+    "chip.qc_report.name": "CEMT0007_MCF10A-H3K4me3"
+}
diff --git a/encode-wrapper/cemt0007_h3k4me3_template.json b/encode-wrapper/cemt0007_h3k4me3_template.json
@@ -0,0 +1,16 @@
+{
+    "chip.pipeline_type" : "histone",
+    "chip.genome_tsv" : "{0}/test_genome_database/hg38_local.tsv",
+    "chip.fastqs" : [
+        [["{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A28471-ATCACG.134224.D2B0LACXX.2.1.fastq.gz",
+          "{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A28471-ATCACG.134224.D2B0LACXX.2.2.fastq.gz"]]
+    ],
+    "chip.ctl_fastqs" : [
+        [["{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.1.fastq.gz",
+          "{0}/v2/ihec/test_data/ChIP-Seq.IX1239-A26691-TCATTC.134224.D2B0LACXX.2.2.fastq.gz"]]
+    ],
+    "chip.paired_end" : true,
+    "chip.choose_ctl.always_use_pooled_ctl" : true,
+    "chip.qc_report.name" : "CEMT0007_MCF10A-H3K4me3",
+    "chip.qc_report.desc" : "CEMT0007_MCF10A H3K4me3"
+}