Sentieon/datametrics: Add plots optionally for each quality metric

.github/workflows/test.yml

@@ -606,6 +606,8 @@ jobs:
             path: modules/nf-core/sentieon/bwaindex
           - profile: conda
             path: modules/nf-core/sentieon/bwamem
+          - profile: conda
+            path: modules/nf-core/sentieon/datametrics
           - profile: conda
             path: modules/nf-core/sentieon/dedup
           - profile: conda

modules/nf-core/sentieon/datametrics/main.nf

@@ -12,6 +12,7 @@ process SENTIEON_DATAMETRICS {
     tuple val(meta), path(bam), path(bai)
     tuple val(meta2), path(fasta)
     tuple val(meta3), path(fai)
+    val plot_results
 
     output:
     tuple val(meta), path('*mq_metrics.txt') , emit: mq_metrics
@@ -20,6 +21,10 @@ process SENTIEON_DATAMETRICS {
     tuple val(meta), path('*gc_metrics.txt') , emit: gc_metrics
     tuple val(meta), path('*aln_metrics.txt'), emit: aln_metrics
     tuple val(meta), path('*is_metrics.txt') , emit: is_metrics
+    tuple val(meta), path('*mq_metrics.pdf') , emit: mq_plot, optional: true
+    tuple val(meta), path('*qd_metrics.pdf') , emit: qd_plot, optional: true
+    tuple val(meta), path('*is_metrics.pdf') , emit: is_plot, optional: true
+    tuple val(meta), path('*gc_metrics.pdf') , emit: gc_plot, optional: true
     path  "versions.yml"                     , emit: versions
 
     when:
@@ -47,6 +52,14 @@ process SENTIEON_DATAMETRICS {
         --algo InsertSizeMetricAlgo ${prefix}_is_metrics.txt  \\
         --algo AlignmentStat ${prefix}_aln_metrics.txt
 
+    if $plot_results
+    then
+        sentieon plot GCBias -o ${prefix}_gc_metrics.pdf ${prefix}_gc_metrics.txt
+        sentieon plot MeanQualityByCycle -o ${prefix}_mq_metrics.pdf ${prefix}_mq_metrics.txt
+        sentieon plot QualDistribution -o ${prefix}_qd_metrics.pdf  ${prefix}_qd_metrics.txt
+        sentieon plot InsertSizeMetricAlgo -o ${prefix}_is_metrics.pdf ${prefix}_is_metrics.txt
+    fi
+
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         sentieon: \$(echo \$(sentieon driver --version 2>&1) | sed -e "s/sentieon-genomics-//g")
@@ -63,6 +76,14 @@ process SENTIEON_DATAMETRICS {
     touch ${prefix}_aln_metrics.txt
     touch ${prefix}_is_metrics.txt
 
+    if $plot_results
+    then
+        touch ${prefix}_gc_metrics.pdf
+        touch ${prefix}_mq_metrics.pdf
+        touch ${prefix}_qd_metrics.pdf
+        touch ${prefix}_is_metrics.pdf
+    fi
+
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         sentieon: \$(echo \$(sentieon driver --version 2>&1) | sed -e "s/sentieon-genomics-//g")

modules/nf-core/sentieon/datametrics/meta.yml

@@ -44,6 +44,10 @@ input:
         type: file
         description: Index of the genome fasta file
         pattern: "*.fai"
+  - - plot_results:
+        type: boolean
+        description: Boolean to determine whether plots should be generated
+        pattern: "true or false"
 output:
   - mq_metrics:
       - meta:
@@ -110,6 +114,46 @@ output:
           description: File containing the information about statistical distribution
             of insert sizes
           pattern: "*.txt"
+  - mq_plot:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'test', single_end:false ]`
+      - "*mq_metrics.pdf":
+          type: file
+          description: "PDF containting plot of mean base quality scores"
+          pattern: "*.pdf"
+  - qd_plot:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'test', single_end:false ]`
+      - "*qd_metrics.pdf":
+          type: file
+          description: "PDF containting plot of specific base quality score"
+          pattern: "*.pdf"
+  - is_plot:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'test', single_end:false ]`
+      - "*is_metrics.pdf":
+          type: file
+          description: "PDF containting plot of insert sizes"
+          pattern: "*.pdf"
+  - gc_plot:
+      - meta:
+          type: map
+          description: |
+            Groovy Map containing sample information
+            e.g. `[ id:'test', single_end:false ]`
+      - "*gc_metrics.pdf":
+          type: file
+          description: "PDF containting plot of GC bias"
+          pattern: "*.pdf"
   - versions:
       - versions.yml:
           type: file

modules/nf-core/sentieon/datametrics/tests/main.nf.test

@@ -0,0 +1,109 @@
+nextflow_process {
+
+    name "Test Process SENTIEON_DATAMETRICS"
+    script "../main.nf"
+    process "SENTIEON_DATAMETRICS"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "sentieon"
+    tag "sentieon/datametrics"
+    config "./nextflow.config"
+
+    test("metrics - no plots") {
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam.bai', checkIfExists: true),
+                ]
+                input [1] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true),
+                ]
+                input [2] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true),
+                ]
+                input [3] = false
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("metrics - with plots") {
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam.bai', checkIfExists: true),
+                ]
+                input [1] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true),
+                ]
+                input [2] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true),
+                ]
+                input [3] = true
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+    test("sarscov2 - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                    [ id:'test', single_end:false ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam', checkIfExists: true),
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/bam/test.paired_end.sorted.bam.bai', checkIfExists: true),
+                ]
+                input [1] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true),
+                ]
+                input [2] = [
+                    [id:'genome'],
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true),
+                ]
+                input [3] = true
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}