Merge pull request #6 from BertGalle/vcf-convert

VCF convert + VEP

.github/workflows/ci.yml

@@ -43,4 +43,4 @@ jobs:
         # For example: adding multiple test runs with different parameters
         # Remember that you can parallelise this by using strategy.matrix
         run: |
-          nextflow run ${GITHUB_WORKSPACE} -profile test,docker --outdir ./results
+          nextflow run ${GITHUB_WORKSPACE} -profile test,docker --outdir ./results -stub

README.md

@@ -1,3 +1,4 @@
+![nf-core/exomecnv](docs/images/nfcore-exomecnv_logo.png)
 [![GitHub Actions CI Status](https://github.com/nf-cmgg/exomecnv/actions/workflows/ci.yml/badge.svg)](https://github.com/nf-cmgg/exomecnv/actions/workflows/ci.yml)
 [![GitHub Actions Linting Status](https://github.com/nf-cmgg/exomecnv/actions/workflows/linting.yml/badge.svg)](https://github.com/nf-cmgg/exomecnv/actions/workflows/linting.yml)[![Cite with Zenodo](http://img.shields.io/badge/DOI-10.5281/zenodo.XXXXXXX-1073c8?labelColor=000000)](https://doi.org/10.5281/zenodo.XXXXXXX)
 [![nf-test](https://img.shields.io/badge/unit_tests-nf--test-337ab7.svg)](https://www.nf-test.com)

assets/exomedepth.yaml

@@ -0,0 +1,53 @@
+# Use --skip 1 with this config
+chrom:
+  value: $6
+
+pos:
+  value: $4
+
+id:
+  value: $7
+
+ref:
+  value: "."
+
+alt:
+  value: ~if $2 == deletion <DEL> <DUP>
+
+filter:
+  value: "."
+
+info:
+  - name: SVTYPE
+    value: ~if $2 == deletion DEL DUP
+    number: 1
+    type: string
+    description: Type of structural variant
+  - name: END
+    value: $5
+    type: integer
+    description: End position of structural variant
+  - name: BF
+    value: $8
+    type: float
+    description: Bayes factor
+  - name: NEXONS
+    value: $3
+    type: integer
+    description: Number of exons
+  - name: READSEXP
+    value: $9
+    type: integer
+    description: Expected reads
+  - name: READSOBS
+    value: $10
+    type: integer
+    description: Observed reads
+  - name: READRATIO
+    value: $11
+    type: float
+    description: Ratio observed/expected reads
+  - name: EXONSENS
+    value: $12
+    type: string
+    description: Exons

assets/schema_input.json

@@ -38,6 +38,20 @@
                 "errorMessage": "index of CRAM or BAM file must be provided, cannot contain spaces and must have extension '.crai' or '.bai' or '.csi'",
                 "format": "file-path",
                 "exists": true
+            },
+            "vcf": {
+                "type": "string",
+                "pattern": "^\\S+\\.(vcf|vcf.gz)$",
+                "errorMessage": "VCF is optional but must be provided when only annotating",
+                "format": "file-path",
+                "exists": true
+            },
+            "tbi": {
+                "type": "string",
+                "pattern": "^\\S+\\.(vcf.gz.tbi)$",
+                "errorMessage": "index of VCF file is optional but must be provided when only annotating",
+                "format": "file-path",
+                "exists": true
             }
         },
         "required": ["sample", "pool", "family", "cram"]

conf/base.config

@@ -59,7 +59,5 @@ process {
         errorStrategy = 'retry'
         maxRetries    = 2
     }
-    withName:CUSTOM_DUMPSOFTWAREVERSIONS {
-        cache = false
-    }
+
 }

conf/modules.config

@@ -18,18 +18,6 @@ process {
         saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
     ]
 
-    withName: FASTQC {
-        ext.args = '--quiet'
-    }
-
-    withName: CUSTOM_DUMPSOFTWAREVERSIONS {
-        publishDir = [
-            path: { "${params.outdir}/pipeline_info" },
-            mode: params.publish_dir_mode,
-            pattern: '*_versions.yml'
-        ]
-    }
-
     withName: 'MULTIQC' {
         ext.args   = { params.multiqc_title ? "--title \"$params.multiqc_title\"" : '' }
         publishDir = [
@@ -39,11 +27,23 @@ process {
         ]
     }
 
-    withName: 'CRAM_TO_BAM' {
-        publishDir = [
-            path: { "${params.outdir}/bam" },
-            mode: params.publish_dir_mode,
-            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
-        ]
+    withName: 'BEDGOVCF' {
+        ext.args = { "--skip 1" }
+    }
+
+    withName: 'ENSEMBLVEP_VEP' {
+        ext.prefix = {"${meta.id}.vep"}
+        ext.args = {[
+            '--vcf',
+            '--format vcf',
+            '--coding_only',
+            '--offline',
+            '--no_stats',
+            '--per_gene',
+            '--merged',
+            '--exclude_predicted',
+            '--fields "SYMBOL,Gene"',
+            '--transcript_filter "stable_id match ENST"'
+        ].join(' ').trim()}
     }
 }

conf/test.config

@@ -24,13 +24,24 @@ params {
     input  = "${projectDir}/assets/samplesheet.csv"
 
     // Genome references
+    // vep_fasta = "/home/bertg/vep/hg38.fa"
     fasta = 'https://raw.githubusercontent.com/nf-cmgg/test-datasets/exomecnv/data/genomics/homo_sapiens/genome/seq/GCA_000001405.15_GRCh38_full_plus_hs38d1_analysis_set_chr21_and_chrX_subset.fna'
     fai = 'https://raw.githubusercontent.com/nf-cmgg/test-datasets/exomecnv/data/genomics/homo_sapiens/genome/seq/GCA_000001405.15_GRCh38_full_plus_hs38d1_analysis_set_chr21_and_chrX_subset.fna.fai'
 
     // Parameters
     roi_auto = "https://raw.githubusercontent.com/nf-cmgg/test-datasets/exomecnv/data/genomics/homo_sapiens/genome/roi/Homo_sapiens.GRCh38.105.chr21_protein_coding_basic_sorted_merged_autosomal.bed"
     roi_chrx = "https://raw.githubusercontent.com/nf-cmgg/test-datasets/exomecnv/data/genomics/homo_sapiens/genome/roi/Homo_sapiens.GRCh38.105.chrX_protein_coding_basic_sorted_merged.bed"
 
-}
+    yamlconfig = "${projectDir}/assets/exomedepth.yaml"
+    exomedepth = false
+
 
+    // VEP options
+    annotate            = false
+    vep_cache           = "path" //Change to correct path
+    vep_assembly        = "GRCh38"
+    vep_cache_version   = 105
+    species             = "homo_sapiens"
+
+}
 

modules.json

@@ -5,14 +5,29 @@
         "https://github.com/nf-core/modules.git": {
             "modules": {
                 "nf-core": {
+                    "bedgovcf": {
+                        "branch": "master",
+                        "git_sha": "71a96063a7e09be7ea48f069d134a56ce86c9ba6",
+                        "installed_by": ["modules"]
+                    },
+                    "ensemblvep/vep": {
+                        "branch": "master",
+                        "git_sha": "b42fec6f7c6e5d0716685cabb825ef6bf6e386b5",
+                        "installed_by": ["modules"]
+                    },
                     "multiqc": {
                         "branch": "master",
                         "git_sha": "b7ebe95761cd389603f9cc0e0dc384c0f663815a",
                         "installed_by": ["modules"]
                     },
                     "samtools/convert": {
                         "branch": "master",
-                        "git_sha": "1b93e06140e6807cc408519759f68bd26da63a00",
+                        "git_sha": "b42fec6f7c6e5d0716685cabb825ef6bf6e386b5",
+                        "installed_by": ["modules"]
+                    },
+                    "tabix/tabix": {
+                        "branch": "master",
+                        "git_sha": "9502adb23c0b97ed8e616bbbdfa73b4585aec9a1",
                         "installed_by": ["modules"]
                     }
                 }
@@ -26,7 +41,7 @@
                     },
                     "utils_nfcore_pipeline": {
                         "branch": "master",
-                        "git_sha": "5caf7640a9ef1d18d765d55339be751bb0969dfa",
+                        "git_sha": "92de218a329bfc9a9033116eb5f65fd270e72ba3",
                         "installed_by": ["subworkflows"]
                     },
                     "utils_nfvalidation_plugin": {

modules/local/exomedepth/cnv_call/environment.yml

@@ -0,0 +1,10 @@
+name: exomedepth_cnv_call
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+
+dependencies:
+  - r=3.6=r36_1003
+  - r-optparse=1.6.4=r36h6115d3f_0
+  - r-exomedepth:1.1.16--r43hfb3cda0_3

modules/local/exomedepth/cnv_call/main.nf

@@ -1,6 +1,8 @@
 process CNV_CALL {
     tag "$sample $meta2.chr"
+    label 'process_low'
 
+    conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
         'https://depot.galaxyproject.org/singularity/r-exomedepth:1.1.16--r43hfb3cda0_3' :
         'biocontainers/r-exomedepth:1.1.16--r43hfb3cda0_3' }"
@@ -12,7 +14,7 @@ process CNV_CALL {
     tuple val(meta2), val(sample), path(countfile) // meta:id, chr, sam, fam, sample
 
     output:
-    tuple val(meta2), path("*.txt"), emit: cnvcall
+    tuple val(meta2), val (sample), path("*.txt"), emit: cnvcall
     path "versions.yml", emit:versions
 
     script:
@@ -34,4 +36,17 @@ process CNV_CALL {
         R: \$(Rscript --version | sed 's/R scripting front-end //g')
     END_VERSIONS
     """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${sample}_CNVs_ExomeDepth_${meta2.chr}"
+    def VERSION = '1.1.16'
+    """
+    touch ${prefix}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        ExomeDepth: ${VERSION}
+        R: \$(Rscript --version | sed 's/R scripting front-end //g')
+    END_VERSIONS
+    """
 }

modules/local/exomedepth/count/environment.yml

@@ -7,4 +7,4 @@ channels:
 dependencies:
   - r=3.6=r36_1003
   - r-optparse=1.6.4=r36h6115d3f_0
-  - r-exomedepth=1.1.12=r36h6786f55_0
+  - r-exomedepth:1.1.16--r43hfb3cda0_3

modules/local/exomedepth/count/main.nf

@@ -39,4 +39,16 @@ process COUNT {
     END_VERSIONS
 
     """
+
+    stub:
+    def VERSION = '1.1.16'
+    """
+    touch ${meta.id}_${meta2.id}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        ExomeDepth: ${VERSION}
+        R: \$(Rscript --version | sed 's/R scripting front-end //g')
+    END_VERSIONS
+    """
 }

modules/local/exomedepth/merge_cnv/environment.yml

@@ -0,0 +1,7 @@
+name: coreutils
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - anaconda::coreutils=9.3

modules/local/exomedepth/merge_cnv/main.nf

@@ -0,0 +1,43 @@
+//MERGE CNV CALL FILES
+process CNV_MERGE {
+    tag "$meta"
+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/coreutils:9.3' :
+        'biocontainers/coreutils:9.3' }"
+    conda "${moduleDir}/environment.yml"
+
+    publishDir "$params.outdir/exomedepth/cnv_call", mode: 'copy'
+
+    input:
+    tuple val(meta), path(auto), path(chrx)
+
+    output:
+    tuple val(meta), path("*.txt"), emit:merge
+    path "versions.yml", emit:versions
+
+    script:
+    def prefix = task.ext.prefix ?: "${meta}_CNVs_ExomeDepth"
+    """
+    cp $auto "${prefix}.txt"
+    tail +2 $chrx >> "${prefix}.txt"
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        tail: \$(tail --version | sed '1!d; s/tail (GNU coreutils) //')
+        cp: \$(cp --version | sed '1!d; s/cp (GNU coreutils) //')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta}_CNVs_ExomeDepth"
+    """
+    touch ${prefix}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        tail: \$(tail --version | sed '1!d; s/tail (GNU coreutils) //')
+        cp: \$(cp --version | sed '1!d; s/cp (GNU coreutils) //')
+    END_VERSIONS
+    """
+}

modules/local/exomedepth/merge_count/environment.yml

@@ -0,0 +1,7 @@
+name: coreutils
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - anaconda::coreutils=9.3

modules/local/exomedepth/merge_count/main.nf

@@ -1,15 +1,23 @@
 // MERGE COUNT FILES
 process COUNT_MERGE {
+    tag "$meta.id $meta.chr"
+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/coreutils:9.3' :
+        'biocontainers/coreutils:9.3' }"
+    conda "${moduleDir}/environment.yml"
+
     publishDir "$params.outdir/exomedepth/counts", mode: 'copy'
 
     input:
     tuple val(meta), path(files)
 
     output:
-    tuple val(meta), path("${prefix}.txt")
+    tuple val(meta), path("*.txt"), emit:merge
+    path "versions.yml", emit:versions
 
     script:
-    prefix = task.ext.prefix ?: "${meta.id}_${meta.chr}"
+    def prefix = task.ext.prefix ?: "${meta.id}_${meta.chr}"
     """
     for file in $files; do
         if [ -f ${prefix}.txt ]; then
@@ -19,5 +27,27 @@ process COUNT_MERGE {
             cp \$file ${prefix}.txt
         fi
     done
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        paste: \$(paste --version | sed '1!d; s/paste (GNU coreutils) //')
+        mv: \$(mv --version | sed '1!d; s/mv (GNU coreutils) //')
+        cp: \$(cp --version | sed '1!d; s/cp (GNU coreutils) //')
+        awk: \$(awk -version 2>&1 | sed '1!d; s/(.*//')
+    END_VERSIONS
+    """
+
+    stub:
+    prefix = task.ext.prefix ?: "${meta.id}_${meta.chr}"
+    """
+    touch ${prefix}.txt
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        paste: \$(paste --version | sed '1!d; s/paste (GNU coreutils) //')
+        mv: \$(mv --version | sed '1!d; s/mv (GNU coreutils) //')
+        cp: \$(cp --version | sed '1!d; s/cp (GNU coreutils) //')
+        awk: \$(awk -version 2>&1 | sed '1!d; s/(.*//')
+    END_VERSIONS
     """
 }