Merge pull request #63 from icgc-argo/dna-seq-alignment@1.8.0

[release]

Author: junjun-zhang

dna-seq-alignment/main.nf

@@ -1,7 +1,7 @@
 #!/usr/bin/env nextflow
 nextflow.enable.dsl = 2
 name = 'dna-seq-alignment'
-version = '1.7.0'
+version = '1.8.0'
 
 /*
 ========================================================================================
@@ -235,19 +235,19 @@ gatkCollectOxogMetrics_params = [
 
 
 // Include all modules and pass params
-include { SongScoreDownload as dnld } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-download@2.6.1/main.nf' params(download_params)
+include { SongScoreDownload as dnld } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-download@2.6.2/main.nf' params(download_params)
 include { seqDataToLaneBam as toLaneBam } from "./modules/raw.githubusercontent.com/icgc-argo/dna-seq-processing-tools/seq-data-to-lane-bam.0.3.3.0/tools/seq-data-to-lane-bam/seq-data-to-lane-bam.nf" params(seqDataToLaneBam_params)
 include {bwaMemAligner; getBwaSecondaryFiles} from "./modules/raw.githubusercontent.com/icgc-argo/dna-seq-processing-tools/bwa-mem-aligner.0.1.12.0/tools/bwa-mem-aligner/bwa-mem-aligner.nf" params(bwaMemAligner_params)
 include { readGroupUBamQC as rgQC } from "./modules/raw.githubusercontent.com/icgc-argo/data-qc-tools-and-wfs/read-group-ubam-qc.0.1.2.0/tools/read-group-ubam-qc/read-group-ubam-qc.nf" params(readGroupUBamQC_params)
 include { bamMergeSortMarkdup as merSorMkdup; getMdupSecondaryFile } from "./modules/raw.githubusercontent.com/icgc-argo/dna-seq-processing-tools/bam-merge-sort-markdup.0.1.11.0/tools/bam-merge-sort-markdup/bam-merge-sort-markdup.nf" params(bamMergeSortMarkdup_params)
 include { alignedSeqQC; getAlignedQCSecondaryFiles } from "./modules/raw.githubusercontent.com/icgc-argo/data-qc-tools-and-wfs/aligned-seq-qc.0.2.2.1/tools/aligned-seq-qc/aligned-seq-qc" params(alignedSeqQC_params)
-include { payloadGenDnaAlignment as pGenDnaAln } from "./modules/raw.githubusercontent.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment.0.3.3.0/tools/payload-gen-dna-alignment/payload-gen-dna-alignment.nf" params(payloadGenDnaAlignment_params)
-include { payloadGenDnaSeqQc as pGenDnaSeqQc } from "./modules/raw.githubusercontent.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-seq-qc.0.5.3.0/tools/payload-gen-dna-seq-qc/payload-gen-dna-seq-qc.nf" params(payloadGenDnaSeqQc_params)
+include { payloadGenDnaAlignment as pGenDnaAln } from "./wfpr_modules/github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment@0.4.0/main.nf" params(payloadGenDnaAlignment_params)
+include { payloadGenDnaSeqQc as pGenDnaSeqQc } from "./wfpr_modules/github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-seq-qc@0.6.0/main.nf" params(payloadGenDnaSeqQc_params)
 include { gatkSplitIntervals as splitItvls; getSecondaryFiles as getSIIdx } from "./modules/raw.githubusercontent.com/icgc-argo/gatk-tools/gatk-split-intervals.4.1.4.1-1.0/tools/gatk-split-intervals/gatk-split-intervals"
 include { metadataParser as mParser } from "./modules/raw.githubusercontent.com/icgc-argo/data-processing-utility-tools/metadata-parser.0.1.0.0/tools/metadata-parser/metadata-parser.nf"
 include { gatkCollectOxogMetrics as oxog; getOxogSecondaryFiles; gatherOxogMetrics as gatherOM } from "./modules/raw.githubusercontent.com/icgc-argo/gatk-tools/gatk-collect-oxog-metrics.4.1.8.0-3.0/tools/gatk-collect-oxog-metrics/gatk-collect-oxog-metrics" params(gatkCollectOxogMetrics_params)
-include { SongScoreUpload as upAln } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.0/main.nf' params(uploadAlignment_params)
-include { SongScoreUpload as upQc } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.0/main.nf' params(uploadQc_params)
+include { SongScoreUpload as upAln } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.1/main.nf' params(uploadAlignment_params)
+include { SongScoreUpload as upQc } from './wfpr_modules/github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.1/main.nf' params(uploadQc_params)
 include { cleanupWorkdir as cleanup } from './modules/raw.githubusercontent.com/icgc-argo/nextflow-data-processing-utility-tools/2.3.0/process/cleanup-workdir'
 include { payloadGenSeqExperiment as pGenExp } from './wfpr_modules/github.com/icgc-argo/data-processing-utility-tools/payload-gen-seq-experiment@0.4.0/main.nf'
 

dna-seq-alignment/pkg.json

@@ -1,6 +1,6 @@
 {
     "name": "dna-seq-alignment",
-    "version": "1.7.0",
+    "version": "1.8.0",
     "description": "ICGC ARGO DNA Seq Processing Workflow",
     "main": "main.nf",
     "deprecated": false,
@@ -15,8 +15,10 @@
     },
     "dependencies": [
         "github.com/icgc-argo/data-processing-utility-tools/payload-gen-seq-experiment@0.4.0",
-        "github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-download@2.6.1",
-        "github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.0"
+        "github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment@0.4.0",
+        "github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-seq-qc@0.6.0",
+        "github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-download@2.6.2",
+        "github.com/icgc-argo/nextflow-data-processing-utility-tools/song-score-upload@2.6.1"
     ],
     "devDependencies": [],
     "contributors": [

modules/raw.githubusercontent.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment.0.3.3.0/tools/payload-gen-dna-alignment/payload-gen-dna-alignment.nf

@@ -1,6 +1,9 @@
 {
   "analysisId": "3e46fd12-a5fc-4d72-86fd-12a5fcbd7216",
   "studyId": "TEST-PR",
+  "info": {
+    "origin": "ICGC_25K"
+  },
   "analysisState": "PUBLISHED",
   "samples": [
     {

tests/data/ee46fd12-a5fc-4d72-86fd-12a5fcbd7216.sequencing_experiment.json

@@ -0,0 +1,5 @@
+.gitignore
+.nextflow*
+tests
+work
+outdir

wfpr_modules/github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment@0.4.0/.dockerignore

@@ -0,0 +1,14 @@
+FROM quay.io/icgc-argo/dna-seq-processing-tools:base-docker.0.2.1
+
+LABEL org.opencontainers.image.source https://github.com/icgc-argo/data-processing-utility-tools
+
+ENV PATH="/tools:${PATH}"
+
+COPY *.py /tools/
+
+WORKDIR /tools
+
+USER ubuntu
+
+ENTRYPOINT ["/usr/bin/env"]
+CMD ["/bin/bash"]

wfpr_modules/github.com/icgc-argo/data-processing-utility-tools/payload-gen-dna-alignment@0.4.0/Dockerfile

@@ -0,0 +1,97 @@
+#!/usr/bin/env nextflow
+
+/*
+  Copyright (C) 2021,  Ontario Institute for Cancer Research
+
+  This program is free software: you can redistribute it and/or modify
+  it under the terms of the GNU Affero General Public License as published by
+  the Free Software Foundation, either version 3 of the License, or
+  (at your option) any later version.
+
+  This program is distributed in the hope that it will be useful,
+  but WITHOUT ANY WARRANTY; without even the implied warranty of
+  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+  GNU Affero General Public License for more details.
+
+  You should have received a copy of the GNU Affero General Public License
+  along with this program.  If not, see <http://www.gnu.org/licenses/>.
+
+  Authors:
+    Linda Xiang
+    Junjun Zhang
+*/
+
+/********************************************************************/
+/* this block is auto-generated based on info from pkg.json where   */
+/* changes can be made if needed, do NOT modify this block manually */
+nextflow.enable.dsl = 2
+version = '0.4.0'  // package version
+
+container = [
+    'ghcr.io': 'ghcr.io/icgc-argo/data-processing-utility-tools.payload-gen-dna-alignment'
+]
+default_container_registry = 'ghcr.io'
+/********************************************************************/
+
+
+// universal params go here
+params.container_registry = ""
+params.container_version = ""
+params.container = ""
+
+params.cpus = 1
+params.mem = 1  // GB
+params.publish_dir = ""  // set to empty string will disable publishDir
+
+
+// tool specific parmas go here, add / change as needed
+params.files_to_upload = ""
+params.seq_experiment_analysis = ""
+params.read_group_ubam_analysis = ""
+params.wf_name = ""
+params.wf_version = ""
+
+
+process payloadGenDnaAlignment {
+  container "${params.container ?: container[params.container_registry ?: default_container_registry]}:${params.container_version ?: version}"
+  publishDir "${params.publish_dir}/${task.process.replaceAll(':', '_')}", mode: "copy", enabled: params.publish_dir
+
+  cpus params.cpus
+  memory "${params.mem} GB"
+
+  input:
+    path files_to_upload
+    path seq_experiment_analysis
+    path read_group_ubam_analysis
+    val wf_name
+    val wf_version
+
+  output:
+    path "*.dna_alignment.payload.json", emit: payload
+    path "out/*", emit: alignment_files
+
+  script:
+    args_read_group_ubam_analysis = read_group_ubam_analysis.size() > 0 ? "-u ${read_group_ubam_analysis}" : ""
+    """
+    main.py \
+      -f ${files_to_upload} \
+      -a ${seq_experiment_analysis} \
+      -w "${wf_name}" \
+      -r ${workflow.runName} \
+      -s ${workflow.sessionId} \
+      -v ${wf_version} ${args_read_group_ubam_analysis}
+    """
+}
+
+
+// this provides an entry point for this main script, so it can be run directly without clone the repo
+// using this command: nextflow run <git_acc>/<repo>/<pkg_name>/<main_script>.nf -r <pkg_name>.v<pkg_version> --params-file xxx
+workflow {
+  payloadGenDnaAlignment(
+    Channel.fromPath(params.files_to_upload).collect(),
+    file(params.seq_experiment_analysis),
+    Channel.fromPath(params.read_group_ubam_analysis).collect(),
+    params.wf_name,
+    params.wf_version
+  )
+}