Merge pull request #624 from umccr/add-qlims-csv-to-bclconvert-interop-qc-pipeline

Author: alexiswl

tools/custom-convert-multiqc-json-data-to-qlims-import-csv/1.0.0/custom-convert-multiqc-json-data-to-qlims-import-csv__1.0.0.cwl

@@ -0,0 +1,92 @@
+cwlVersion: v1.1
+class: CommandLineTool
+
+# Extensions
+$namespaces:
+    s: https://schema.org/
+    ilmn-tes: https://platform.illumina.com/rdf/ica/
+$schemas:
+  - https://schema.org/version/latest/schemaorg-current-http.rdf
+
+# Metadata
+s:author:
+    class: s:Person
+    s:name: Alexis Lucattini
+    s:email: [email protected]
+    s:identifier: https://orcid.org/0000-0001-9754-647X
+
+# ID/Docs
+id: custom-convert-multiqc-json-data-to-qlims-import-csv--1.0.0
+label: custom-convert-multiqc-json-data-to-qlims-import-csv v(1.0.0)
+doc: |
+    Documentation for custom-convert-multiqc-json-data-to-qlims-
+    import-csv v1.0.0
+
+# ILMN V1 Resources Guide: https://illumina.gitbook.io/ica-v1/analysis/a-taskexecution#type-and-size
+# ILMN V2 Resources Guide: https://help.ica.illumina.com/project/p-flow/f-pipelines#compute-types
+hints:
+    ResourceRequirement:
+        ilmn-tes:resources/tier: standard
+        ilmn-tes:resources/type: standard
+        ilmn-tes:resources/size: small
+        coresMin: 2
+        ramMin: 4000
+    DockerRequirement:
+        dockerPull: ghcr.io/umccr/alpine-jq:1.6
+
+requirements:
+  InlineJavascriptRequirement: {}
+  InitialWorkDirRequirement:
+    listing:
+      - entryname: "convert-multiqc-json-data-to-qlims-import-csv.sh"
+        entry: |
+          jq --raw-output \\
+            --argjson human_genome_size "$(inputs.human_genome_size)" \\
+            '
+              .report_saved_raw_data.multiqc_bclconvert_bysample | to_entries |
+              map(
+                .key as $sample_name |
+                .value.lanes | to_entries |
+                map(
+                  .key as $lane |
+                  {
+                    "Sample Name": $sample_name,
+                    "Coverage": ((.value.yield_q30 / $human_genome_size) * 100 | round / 100),
+                    "Lane": ($lane[1:] | tonumber),
+                    "Clusters": .value.clusters,
+                    "One mismatch index": ((.value.percent_one_mismatch_index_reads) * 100 | round / 100),
+                    "Yield": .value.yield_
+                  }
+                )
+              ) |
+              flatten |
+              # To csv (With ordered keys)
+              # https://unix.stackexchange.com/questions/754935/convert-json-to-csv-with-headers-in-jq
+              [first|keys_unsorted] + map([.[]]) | .[] | @csv |
+              # Remove quotes
+              # https://stackoverflow.com/questions/60350315/why-does-the-jq-raw-output-argument-fail-to-remove-quotes-from-csv-output
+              gsub("\\""; "")
+          ' < "$(inputs.multiqc_data_json.path)"
+
+baseCommand: [ "bash", "convert-multiqc-json-data-to-qlims-import-csv.sh" ]
+
+stdout: "qlims.csv"
+
+inputs:
+  human_genome_size:
+    label: "Human genome size"
+    type: long
+    default: 3049315783
+  multiqc_data_json:
+    label: "MultiQC JSON data"
+    type: File
+
+outputs:
+  qlims_csv:
+    type: File
+    outputBinding:
+      glob: "qlims.csv"
+
+
+successCodes:
+  - 0

workflows/bclconvert-interop-qc/1.3.1--1.25.2/bclconvert-interop-qc__1.3.1--1.25.2.cwl

@@ -122,6 +122,43 @@ steps:
       - id: output_file
     run: ../../../tools/multiqc/1.25.2/multiqc__1.25.2.cwl
 
+  # Generate the qlims csv file
+  generate_qc_csv_step:
+    label: Generate QLIMS CSV
+    doc: |
+      Generate the QLIMS CSV file from the interop directory
+    in:
+      multiqc_data_json:
+        source:
+          - run_multiqc_step/output_directory
+          - instrument_run_id
+        valueFrom: |
+          ${
+            return {
+              "class": "File",
+              "basename": "multiqc_data.json",
+              "location": self[0].location + "/" + self[1] + "_multiqc_report_data/multiqc_data.json"
+            };
+          }
+    out:
+      - id: qlims_csv
+    run: ../../../tools/custom-convert-multiqc-json-data-to-qlims-import-csv/1.0.0/custom-convert-multiqc-json-data-to-qlims-import-csv__1.0.0.cwl
+
+  rename_qc_csv_step:
+    label: Rename QLIMS CSV
+    doc: |
+      Rename the QLIMS CSV file to include the instrument run ID
+    in:
+      basename:
+        source: instrument_run_id
+        valueFrom: "$(self)_multiqc_bclconvert_summary_qlims.csv"
+      location:
+        source: generate_qc_csv_step/qlims_csv
+        valueFrom: "$(self.location)"
+    out:
+      - id: output_file
+    run: ../../../expressions/create-file/1.0.0/create-file__1.0.0.cwl
+
 outputs:
   # Collect everything
   interop_output_dir:
@@ -142,3 +179,12 @@ outputs:
       The HTML report generated by the multiqc step
     type: File
     outputSource: run_multiqc_step/output_file
+  multiqc_qlims_csv_report:
+    label: multiqc qlims csv report
+    doc: |
+      The CSV required by qlims for ingestion generated from the multiqc step
+    type: File
+    outputSource: rename_qc_csv_step/output_file
+
+
+