Merge branch 'master' into master

.github/workflows/main.yml

@@ -16,7 +16,7 @@ jobs:
     needs: test_all_profiles
     strategy:
         matrix:
-          profile: ['abims', 'awsbatch', 'aws_tower', 'bi','bigpurple', 'binac', 'biohpc_gen', 'cbe', 'ccga_dx', 'ccga_med', 'cfc', 'cfc_dev', 'crick', 'denbi_qbic', 'ebc', 'eddie', 'eva', 'genotoul', 'genouest', 'gis', 'google', 'hebbe', 'icr_davros', 'ifb_core', 'imperial', 'imperial_mb', 'jax', 'kraken', 'mpcdf', 'munin', 'oist', 'pasteur', 'phoenix', 'prince', 'sanger', 'seg_globe', 'shh', 'uct_hpc', 'uppmax', 'utd_ganymede', 'uzh']
+          profile: ['abims', 'awsbatch', 'aws_tower', 'bi','bigpurple', 'binac', 'biohpc_gen', 'cbe', 'ccga_dx', 'ccga_med', 'cfc', 'cfc_dev', 'crick', 'denbi_qbic', 'ebc', 'eddie', 'eva', 'genotoul', 'genouest', 'gis', 'google', 'hebbe', 'icr_davros', 'ifb_core', 'imperial', 'imperial_mb', 'jax', 'kraken', 'mpcdf', 'munin', 'oist', 'pasteur', 'phoenix', 'prince', 'sanger', 'seg_globe', 'shh', 'uct_hpc', 'uppmax', 'utd_ganymede', 'utd_sysbio' ,'uzh']
     steps:
       - uses: actions/checkout@v1
       - name: Install Nextflow

README.md

@@ -131,6 +131,7 @@ Currently documentation is available for the following systems:
 * [UCT_HPC](docs/uct_hpc.md)
 * [UPPMAX](docs/uppmax.md)
 * [UTD_GANYMEDE](docs/utd_ganymede.md)
+* [UTD_SYSBIO](docs/utd_sysbio.md)
 * [UZH](docs/uzh.md)
 
 ### Uploading to `nf-core/configs`

conf/pipeline/eager/eva.config

@@ -69,7 +69,7 @@ process {
    }
 
   withName: adapter_removal {
-    clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.toGiga() * 2)}G,virtual_free=${(task.memory.toGiga() * 2)}G" }
+    clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.toGiga() * 4)}G,virtual_free=${(task.memory.toGiga() * 4)}G" }
    }
 
   withName: dedup {
@@ -97,7 +97,7 @@ process {
    }
 
   withName: mtnucratio {
-    clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.toGiga() * 2)}G,virtual_free=${(task.memory.toGiga() * 2)}G" }
+    clusterOptions = { "-S /bin/bash -l h_vmem=${(task.memory.toGiga() * 4)}G,virtual_free=${(task.memory.toGiga() * 4)}G" }
    }
 
   withName: vcf2genome {

conf/pipeline/rnaseq/utd_sysbio.config

@@ -0,0 +1,19 @@
+params {
+  config_profile_description = 'University of Texas at Dallas HPC cluster profile provided by nf-core/configs'
+  config_profile_contact = 'Edmund Miller(@emiller88)'
+  config_profile_url = 'http://docs.oithpc.utdallas.edu/'
+}
+
+process {
+
+  withName : "STAR_ALIGN" {
+    memory = 36.GB
+  }
+
+  withLabel:process_high {
+    cpus = { check_max( 16 * task.attempt, 'cpus' ) }
+    memory = { check_max( 60.GB * task.attempt, 'memory' ) }
+    time = { check_max( 16.h * task.attempt, 'time' ) }
+  }
+
+}

conf/pipeline/viralrecon/genomes.config

@@ -20,23 +20,29 @@ params {
       primer_sets {
         artic {
           '1' {
-            fasta      = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V1/nCoV-2019.reference.fasta'
+            fasta      = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V1/nCoV-2019.reference.fasta'
             gff        = 'https://github.com/nf-core/test-datasets/raw/viralrecon/genome/MN908947.3/GCA_009858895.3_ASM985889v3_genomic.200409.gff.gz'
-            primer_bed = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V1/nCoV-2019.primer.bed'
+            primer_bed = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V1/nCoV-2019.primer.bed'
             scheme     = 'nCoV-2019'
           }
           '2' {
-            fasta      = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V2/nCoV-2019.reference.fasta'
+            fasta      = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V2/nCoV-2019.reference.fasta'
             gff        = 'https://github.com/nf-core/test-datasets/raw/viralrecon/genome/MN908947.3/GCA_009858895.3_ASM985889v3_genomic.200409.gff.gz'
-            primer_bed = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V2/nCoV-2019.primer.bed'
+            primer_bed = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V2/nCoV-2019.primer.bed'
             scheme     = 'nCoV-2019'
           }
           '3' {
-            fasta      = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V3/nCoV-2019.reference.fasta'
+            fasta      = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V3/nCoV-2019.reference.fasta'
             gff        = 'https://github.com/nf-core/test-datasets/raw/viralrecon/genome/MN908947.3/GCA_009858895.3_ASM985889v3_genomic.200409.gff.gz'
-            primer_bed = 'https://github.com/artic-network/primer-schemes/raw/master/nCoV-2019/V3/nCoV-2019.primer.bed'
+            primer_bed = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V3/nCoV-2019.primer.bed'
             scheme     = 'nCoV-2019'
           }
+          '4' {
+            fasta      = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V4/SARS-CoV-2.reference.fasta'
+            gff        = 'https://github.com/nf-core/test-datasets/raw/viralrecon/genome/MN908947.3/GCA_009858895.3_ASM985889v3_genomic.200409.gff.gz'
+            primer_bed = 'https://github.com/artic-network/artic-ncov2019/raw/master/primer_schemes/nCoV-2019/V4/SARS-CoV-2.scheme.bed'
+            scheme     = 'SARS-CoV-2'
+          }
         }
       }
     }

conf/utd_sysbio.config

@@ -0,0 +1,35 @@
+//Profile config names for nf-core/configs
+params {
+  config_profile_description = 'University of Texas at Dallas HPC cluster profile provided by nf-core/configs'
+  config_profile_contact = 'Edmund Miller(@emiller88)'
+  config_profile_url = 'http://docs.oithpc.utdallas.edu/'
+  singularity_cache_dir = '/scratch/applied-genomics/singularity'
+}
+
+env {
+    TMPDIR = '/home/$USER/scratch/tmp'
+}
+
+singularity {
+  enabled = true
+  envWhitelist='SINGULARITY_BINDPATH'
+  autoMounts = true
+  cacheDir = params.singularity_cache_dir
+}
+
+process {
+  beforeScript = 'module load singularity/3.4.1'
+  executor = 'slurm'
+  queue = { task.memory >= 30.GB && task.cpu <= 16 ? 'normal': 'smallmem' }
+}
+
+// Preform work directory cleanup after a successful run
+cleanup = true
+
+params {
+  // TODO Need to initialize this
+  // igenomes_base = '/scratch/applied-genomics/references/iGenomes/references/'
+  max_memory = 90.GB
+  max_cpus = 16
+  max_time = 96.h
+}

docs/pipeline/rnaseq/utd_sysbio.md

@@ -0,0 +1,23 @@
+# nf-core/configs: UTD sysbio rnaseq specific configuration
+
+Extra specific configuration for rnaseq pipeline
+
+## Usage
+
+To use, run the pipeline with `-profile sysbio`.
+
+This will download and launch the eager specific [`utd_sysbio.config`](../../../conf/pipeline/rnaseq/utd_sysbio.config) which has been pre-configured with a setup suitable for the shh cluster.
+
+Example: `nextflow run nf-core/rnaseq -profile sysbio`
+
+## rnaseq specific configurations for UTD sysbio
+
+Specific configurations for UTD Sysbio has been made for rnaseq.
+
+### General profiles
+
+<!-- TODO -->
+
+### Contextual profiles
+
+<!-- TODO -->

docs/utd_sysbio.md

@@ -0,0 +1,16 @@
+# nf-core/configs: UTD Sysbio Configuration
+
+All nf-core pipelines have been successfully configured for use on the Sysbio HPC cluster at the [The Univeristy of Texas at Dallas](https://www.utdallas.edu/).
+
+To use, run the pipeline with `-profile utd_sysbio`. This will download and launch the [`utd_sysbio.config`](../conf/utd_sysbio.config) which has been pre-configured with a setup suitable for the Sysbio HPC cluster. Using this profile, a docker image containing all of the required software will be downloaded, and converted to a Singularity image before execution of the pipeline.
+
+Before running the pipeline you will need to load Singularity using the environment module system on Sysbio. You can do this by issuing the commands below:
+
+```bash
+## Singularity environment modules
+module purge
+module load singularity
+```
+
+>NB: You will need an account to use the HPC cluster on Sysbio in order to run the pipeline. If in doubt contact OIT.
+>NB: Nextflow will need to submit the jobs via SLURM to the HPC cluster and as such the commands above will have to be executed on one of the login nodes. If in doubt contact OIT.

nfcore_custom.config

@@ -50,6 +50,7 @@ profiles {
   uct_hpc      { includeConfig "${params.custom_config_base}/conf/uct_hpc.config" }
   uppmax       { includeConfig "${params.custom_config_base}/conf/uppmax.config" }
   utd_ganymede { includeConfig "${params.custom_config_base}/conf/utd_ganymede.config" }
+  utd_sysbio   { includeConfig "${params.custom_config_base}/conf/utd_sysbio.config" }
   uzh          { includeConfig "${params.custom_config_base}/conf/uzh.config" }
   jax          { includeConfig "${params.custom_config_base}/conf/jax.config" }
 }
@@ -70,6 +71,7 @@ params {
     genotoul: ['.genologin1.toulouse.inra.fr', '.genologin2.toulouse.inra.fr'],
     genouest: ['.genouest.org'],
     uppmax: ['.uppmax.uu.se'],
-    utd_ganymede: ['ganymede.utdallas.edu']
+    utd_ganymede: ['ganymede.utdallas.edu'],
+    utd_sysbio: ['sysbio.utdallas.edu']
   ]
 }

pipeline/rnaseq.config

@@ -10,4 +10,5 @@
 
 profiles {
   eddie { includeConfig "${params.custom_config_base}/conf/pipeline/rnaseq/eddie.config" }
+  utd_sysbio { includeConfig "${params.custom_config_base}/conf/pipeline/rnaseq/utd_sysbio.config" }
 }