Add preprocessing in DSL2

CHANGELOG.md

@@ -7,7 +7,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/) a
 ## [dsl2](https://github.com/nf-core/sarek/tree/dev)
 
 - [#238](https://github.com/nf-core/sarek/pull/238) -Add subworkflow for building all the indices
-
+- [#241](https://github.com/nf-core/sarek/pull/241) -Add modules and workflows parts for preprocessing steps
 ## [dev](https://github.com/nf-core/sarek/tree/dev)
 
 - [#234](https://github.com/nf-core/sarek/pull/234) -Switching to DSL2

assets/multiqc_config.yaml

@@ -19,6 +19,7 @@ top_modules:
      name: 'FastQC'
      path_filters_exclude: 
         - '*trimmed_fastqc*'
+- 'cutadapt'
 - 'fastqc':
      name: 'FastQC after trimming'
      info: 'FastQC after applying TrimGalore.'

bin/scrape_software_versions.py

@@ -7,7 +7,7 @@
     'AlleleCount': ['v_allelecount.txt', r"(\S+)"],
     'ASCAT': ['v_ascat.txt', r"Version:       (\S+)"],
     'bcftools': ['v_bcftools.txt', r"bcftools (\S+)"],
-    'BWA': ['v_bwa.txt', r"Version: (\S+)"],
+    'BWAMEM2': ['v_bwamem2.txt', r"Version: (\S+)"],
     'CNVkit': ['v_cnvkit.txt', r"(\S+)"],
     'Control-FREEC': ['v_controlfreec.txt', r"Control-FREEC\s(\S+)"],
     'FastQC': ['v_fastqc.txt', r"FastQC v(\S+)"],
@@ -35,7 +35,7 @@
 results['ASCAT'] = '<span style="color:#999999;\">N/A</span>'
 results['AlleleCount'] = '<span style="color:#999999;\">N/A</span>'
 results['bcftools'] = '<span style="color:#999999;\">N/A</span>'
-results['BWA'] = '<span style="color:#999999;\">N/A</span>'
+results['BWA-MEM2'] = '<span style="color:#999999;\">N/A</span>'
 results['CNVkit'] = '<span style="color:#999999;\">N/A</span>'
 results['Control-FREEC'] = '<span style="color:#999999;\">N/A</span>'
 results['FastQC'] = '<span style="color:#999999;\">N/A</span>'

conf/base.config

@@ -19,55 +19,55 @@ process {
   maxErrors = '-1'
   maxRetries = 3
 
-  withLabel:cpus_1 {
+  withLabel:CPUS_1 {
     cpus = {check_resource(1)}
   }
-  withLabel:cpus_2 {
+  withLabel:CPUS_2 {
     cpus = {check_resource(2)}
   }
-  withLabel:cpus_4 {
+  withLabel:CPUS_4 {
     cpus = {check_resource(4)}
   }
-  withLabel:cpus_8 {
+  withLabel:CPUS_8 {
     cpus = {check_resource(8)}
   }
-  withLabel:cpus_16 {
+  withLabel:CPUS_16 {
     cpus = {check_resource(16)}
   }
-  withLabel:cpus_max {
+  withLabel:CPUS_MAX {
     cpus = {params.max_cpus}
   }
 
-  withLabel:memory_singleCPU_2_task {
+  withLabel:MEMORY_SINGLECPU_2_TASK {
     memory = {check_resource((params.single_cpu_mem as nextflow.util.MemoryUnit) * 2 * task.attempt)}
   }
-  withLabel:memory_singleCPU_task_sq {
+  withLabel:MEMORY_SINGLECPU_TASK_SQ {
     memory = {check_resource((params.single_cpu_mem as nextflow.util.MemoryUnit) * task.attempt * task.attempt)}
   }
-  withLabel:memory_max {
+  withLabel:MEMORY_MAX {
     memory = {params.max_memory}
   }
 
-  withName:Get_software_versions {
+  withName:GET_SOFTWARE_VERSIONS {
     cache = false
   }
 
-  withName:ConcatVCF {
-    // For unknown reasons, ConcatVCF sometimes fails with SIGPIPE
+  withName:CONCATVCF {
+    // For unknown reasons, CONCATVCF sometimes fails with SIGPIPE
     // (exit code 141). Rerunning the process will usually work.
     errorStrategy = {task.exitStatus == 141 ? 'retry' : 'terminate'}
   }
-  withLabel:FastQC {
+  withLabel:FASTQC {
     errorStrategy = {task.exitStatus == 143 ? 'retry' : 'ignore'}
   }
-  withName:MapReads {
+  withName:BWAMEM2_MEM {
     memory = {check_resource(60.GB * task.attempt)}
     time = {check_resource(48.h * task.attempt)}
   }
-  withName:MultiQC {
+  withName:MULTIQC {
     errorStrategy = {task.exitStatus == 143 ? 'retry' : 'ignore'}
   }
-  withName:Snpeff {
+  withName:SNPEFF {
     container = {(params.annotation_cache && params.snpeff_cache) ? 'nfcore/sarek:dev' : "nfcore/sareksnpeff:dev.${params.genome}"}
     errorStrategy = {task.exitStatus == 143 ? 'retry' : 'ignore'}
   }

lib/Schema.groovy

@@ -112,10 +112,10 @@ class JSON {
 
     static String params_mqc_summary(summary) {
         String yaml_file_text  = """
-        id: 'nf-core-tcrseq-summary'
+        id: 'nf-core-sarek-summary'
         description: " - this information is collected when the pipeline is started."
-        section_name: 'nf-core/tcrseq Workflow Summary'
-        section_href: 'https://github.com/nf-core/tcrseq'
+        section_name: 'nf-core/sarek Workflow Summary'
+        section_href: 'https://github.com/nf-core/sarek'
         plot_type: 'html'
         data: |
             <dl class=\"dl-horizontal\">