diff --git a/.gitignore b/.gitignore
index 7756350a2c67d0e3d37642c65031fe63689233da..0ee8eb31fed6a000d253a18e8c9648776c984624 100644
--- a/.gitignore
+++ b/.gitignore
@@ -102,10 +102,13 @@ ENV/
.DS_Store
# nextflow analysis folders/files
-pipeline_trace*.txt*
-.nextflow*.log*
-report*.html*
-timeline*.html*
-/workflow/output/*
-/work/*
-/.nextflow/*
+workflow/dag*.dot*
+workflow/trace*.txt*
+workflow/.nextflow/
+workflow/.nextflow*.log*
+workflow/report*.html*
+workflow/timeline*.html*
+workflow/output/*
+workflow/work/*
+workflow/images/
+workflow/cleanup.sh
diff --git a/astrocyte_pkg.yml b/astrocyte_pkg.yml
index da8cb77fc6ac9791ead218425c51215927c87d46..bc1eb2c358835658a79c5995f07aef775fe8daf1 100644
--- a/astrocyte_pkg.yml
+++ b/astrocyte_pkg.yml
@@ -7,13 +7,13 @@
# -----------------------------------------------------------------------------
# A unique identifier for the workflow package, text/underscores only
-name: 'atac-seq-source'
+name: 'atac_seq_source'
# Who wrote this?
author: 'Achisha Saikia, Felix Perez, Peng Lian'
# A contact email address for questions
email: 'achisha.saikia@utsouthwestern.edu, felix.perez@utsouthwestern.edu, biohpc-help@utsouthwestern.edu'
# A more informative title for the workflow package
-title: 'ATAC-seq Source Workflow"
+title: 'ATAC-seq Source Workflow'
# A summary of the workflow package in plain text
description: |
# TODO: Please describe the workflow. (AS)
@@ -24,19 +24,24 @@ citation: |
used, and the overall pipeline doi: 12.3456/zenodo.9876543.
# TODO: Review "individual programs".
Please cite in publications:
- The oringal ATAC-seq pipeline was developed by the ENCODE team, and adapted to the Astrocyte platform by a collaboration between the Lee Lab in Bioinformatics at UT Southwestern Medical Center & BioHPC.
-
-# The minimum Astrocyte version that requires to run this workflow. For old pipelines, which do not have this label
+ The oringal ATAC-seq pipeline was developed by the ENCODE team, and
+ adapted to the Astrocyte Workflow platform by a collaboration
+ at UT Southwestern Medical Center between
+ the Lee (Jeon) Lab in Bioinformatics & BioHPC.
+
+# The minimum Astrocyte version that requires to run this workflow. For old pipelines, which do not have this label
# a default value of 0.3.1 will be assigned automatically. A request of minimum version less than 0.4.0 will be ignored.
minimum_astrocyte_version: '2.0.1'
-# The Nextflow version that requires to run this workflow. For old pipelines, which do not have this label
+# The Nextflow version that requires to run this workflow. For old pipelines, which do not have this label
# a default value of 0.31.0 will be assigned automatically. Please make sure the requested nextflow version is available
# in the module list.
-nextflow_version: '22.04.5'
+nextflow_version: '0.31.0'
# (Optional) The Nextflow config file to use for this workflow. If provided, the file should exist in workflow/configs
nextflow_config: 'biohpc.config'
# The container to use for this workflow, none/singularity. If omitted, the default value 'none' will be used.
-container: 'none'
+container: 'singularity'
+# The version of singularity to use. This is required if container == 'singularity'
+singularity_version: '3.9.9'
# -----------------------------------------------------------------------------
@@ -63,6 +68,15 @@ documentation_files:
workflow_modules:
- Test
+# A list of container images required to run this workflow.
+# Specify full path and version names to ensure reproducibility.
+# This keyword is required when 'container' is specified in Astrocyte 0.4.1 and above.
+# Singularity supports different registries, please specify the protocol to use.
+# Such as, "docker://", "shub://", "library://", etc. We encourage you to use the GitLab
+# container registry of BioHPC to save and manage your container images.
+workflow_containers:
+ - docker://git.biohpc.swmed.edu:5050/s219741/astrocyte-atac-source/atac:0.0.1
+
# A list of parameters used by the workflow, defining how to present them,
# options etc in the web interface. For each parameter:
#
@@ -88,7 +102,7 @@ workflow_modules:
# SELECT TYPE
# choices: A set of choices presented to the user for the parameter.
# Each choice is a pair of value and description, e.g.
-#
+#
# choices:
# - [ 'myval', 'The first option']
# - [ 'myval', 'The second option']
@@ -99,25 +113,11 @@ workflow_modules:
workflow_parameters:
- id: input_json
- # START HERE NEXT TIME
- type: string
- required: false
- default: "A"
+ type: file
+ required: true
description: |
- A parameter that need to be passed to the above command
+ The input JSON used to supply the list of fastq.gz files for caper to process in the ATAC-seq pipeline.
- - id: parameter2
- type: string
- required: false
- default: "B"
- description: |
- A parameter that need to be passed to the above command
-
- - id: parameter_nonopt
- type: files
- required: false
- description: |
- A non-optinal parameter that need to be passed to the above command
# -----------------------------------------------------------------------------
# SHINY APP CONFIGURATION
@@ -128,22 +128,21 @@ workflow_parameters:
# Name of the R module that the vizapp will run against
# A containerized version is recommanded.
-vizapp_r_module: 'R/4.1.1-img'
+#vizapp_r_module: 'R/4.1.1-img'
# List of any CRAN packages, not provided by the modules, that must be made
# available to the vizapp
-vizapp_cran_packages:
- - shiny
- - shinyFiles
- - plotly
+#vizapp_cran_packages:
+# - shiny
+# - shinyFiles
+# - plotly
# # List of any Bioconductor packages, not provided by the modules, that must be made
# available to the vizapp
-vizapp_bioc_packages:
- -
+#vizapp_bioc_packages:
+# -
# # List of any packages to install from GitHub using devtools, that must be
# made availavle to the vizapp
-vizapp_github_packages:
- -
-
+#vizapp_github_packages:
+# -
diff --git a/workflow/configs/biohpc.config b/workflow/configs/biohpc.config
old mode 100644
new mode 100755
index 4f2921ae0e59bf5aa79168ae463e3a6c0227fae5..465d7e7bd09bbd7b6bcf8541c74fc87b271d0b73
--- a/workflow/configs/biohpc.config
+++ b/workflow/configs/biohpc.config
@@ -1,9 +1,22 @@
+singularity {
+ enabled = true
+ runOptions = ' -B /cm/shared/apps/slurm/16.05.8 -B /etc/slurm -B /cm/shared/apps/slurm/var/etc/ -B /usr/lib64/libreadline.so.6 -B /usr/lib64/libhistory.so.6 -B /usr/lib64/libtinfo.so.5 -B /var/run/munge -B /usr/lib64/libmunge.so.2 -B /usr/lib64/libmunge.so.2.0.0 -B /cm/shared/apps/slurm/16.05.8/lib64/slurm/ -B /cm/shared/apps/slurm/16.05.8/lib64'
+ // Please do NOT use "--disable-cache" in this runOptions.
+ // Starting from version 2.0.0, the astrocyte_cli will clean up the cache automatically.
+ // runOptions = '--bind /vagrant:/vagrant' // Use this one for vagrant development env only
+ cacheDir = "$baseDir/images/singularity" // Singularity images specified in `workflow_containers` of astrocyte_pkg.yml will be saved to
+ // this folder automatically, before running the workflow. The images will be renamed as
+ // "NAME-TAG.img", e.g. ubuntu-latest.img, centos-centos8.img, r-4.1.1.img, etc.
+}
+
process {
- // Uncomment these two options to run the workflow on BioHPC cluster
- // You can
- // executor = 'slurm'
- // queue = 'super'
+ executor = 'slurm'
+ clusterOptions = '--hold --no-kill'
+ queue = 'super'
+ beforeScript = 'ulimit -Ss unlimited'
- // To run the workflow on local computer (for testing)
- executor = 'local'
+ withName:runSource {
+ container = 'docker://git.biohpc.swmed.edu:5050/s219741/astrocyte-atac-source/atac:0.0.1'
+ executor = 'local'
+ }
}
diff --git a/workflow/main.nf b/workflow/main.nf
index 25f81fe3d0b0420c385dbadde44baf0404daba65..734b08d987be1c6728308c943e3c79f21d14b27a 100644
--- a/workflow/main.nf
+++ b/workflow/main.nf
@@ -1,15 +1,32 @@
/*
* Copyright (c) 2024. The University of Texas Southwestern Medical Center
*
- * TODO: (AC) Brief description of ATAC-seq
+ * TODO: (AS) Brief description of ATAC-seq
*
* @authors
* Achisha Saikia, Felix Perez
*
*/
- // caper hpc submit atac.wdl -i "${INPUT_JSON}" --singularity --leader-job-name ANY_GOOD_LEADER_JOB_NAME`
+// Note - $baseDir is the location of this workflow's main.nf file.
+params.input_json = "$baseDir/../test_data/example_input_json/ENCSR356KRQ_subsampled.json"
- // The executable script in the Runner repo
- //params = "caper hpc submit $baseDir/external_repo/atac.wdl"
- //params.input_json =
\ No newline at end of file
+inputJsonFile = Channel.fromPath( params.input_json )
+
+process runSource {
+
+ publishDir "$baseDir/output", mode: 'copy'
+
+ input:
+ file inputJson from inputJsonFile
+
+ output:
+ file '*'
+
+ """
+ export LD_LIBRARY_PATH=/usr/lib64/:$LD_LIBRARY_PATH
+ sinfo -V > slurm_version.txt
+ caper --version > caper_version.txt
+ caper hpc submit $baseDir/external_repo/astrocyte-atac-runner/atac.wdl -i $inputJson --singularity --leader-job-name atac-source 1> batch_job.txt 2>> caper_err.txt
+ """
+}
\ No newline at end of file