Update lofreq somatic and callparallel (#5931)

* update lofreq callparallel and somatic ssh * [automated] Fix linting with Prettier * suggestions * remove files tests * label lofreq somatic * Update modules/nf-core/lofreq/callparallel/main.nf Co-authored-by: Famke Bäuerle <[email protected]> * port lofreq/callparallel to nf-test * Swap to nftest * update paths * add meta * Update main.nf * modify main * update paths * fix small errors * Update modules/nf-core/lofreq/somatic/tests/main.nf.test Co-authored-by: Simon Pearce <[email protected]> * Update main.nf.test.snap --------- Co-authored-by: AitorOP <[email protected]> Co-authored-by: nf-core-bot <[email protected]> Co-authored-by: Famke Bäuerle <[email protected]> Co-authored-by: Francisco Martínez <[email protected]> Co-authored-by: Famke Bäuerle <[email protected]> Co-authored-by: Simon Pearce <[email protected]>
nf-core · Jul 11, 2024 · 94e871f · 94e871f
1 parent 882e20c
commit 94e871f
Show file tree

Hide file tree

Showing 16 changed files with 379 additions and 156 deletions.
diff --git a/modules/nf-core/lofreq/callparallel/main.nf b/modules/nf-core/lofreq/callparallel/main.nf
@@ -8,9 +8,9 @@ process LOFREQ_CALLPARALLEL {
         'biocontainers/lofreq:2.1.5--py38h588ecb2_4' }"
 
     input:
-    tuple val(meta), path(bam), path(bai), path(intervals)
-    path fasta
-    path fai
+    tuple val(meta) , path(bam), path(bai), path(intervals)
+    tuple val(meta2), path(fasta)
+    tuple val(meta3), path(fai)
 
     output:
     tuple val(meta), path("*.vcf.gz"), emit: vcf
@@ -23,15 +23,42 @@ process LOFREQ_CALLPARALLEL {
     def args = task.ext.args ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
     def options_intervals = intervals ? "-l ${intervals}" : ""
+
+    def alignment_cram =  bam.Extension == "cram" ? true : false
+    def alignment_bam = bam.Extension == "bam" ? true : false
+    def alignment_out = alignment_cram ? bam.BaseName + ".bam" : "${bam}"
+
+    def samtools_cram_convert = ''
+    samtools_cram_convert += alignment_cram ? "    samtools view -T ${fasta} ${bam} -@ $task.cpus -o ${alignment_out}\n" : ''
+    samtools_cram_convert += alignment_cram ? "    samtools index ${alignment_out}\n" : ''
+
+    def samtools_cram_remove = ''
+    samtools_cram_remove += alignment_cram ? "     rm ${alignment_out}\n" : ''
+    samtools_cram_remove += alignment_cram ? "     rm ${alignment_out}.bai\n " : ''
     """
+    $samtools_cram_convert
+
     lofreq \\
         call-parallel \\
         --pp-threads $task.cpus \\
         $args \\
         $options_intervals \\
         -f $fasta \\
         -o ${prefix}.vcf.gz \\
-        $bam
+        $alignment_out
+
+    $samtools_cram_remove
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        lofreq: \$(echo \$(lofreq version 2>&1) | sed 's/^version: //; s/ *commit.*\$//')
+    END_VERSIONS
+    """
+
+    stub:
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    echo "" | gzip > ${prefix}.vcf.gz
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

diff --git a/modules/nf-core/lofreq/callparallel/meta.yml b/modules/nf-core/lofreq/callparallel/meta.yml
@@ -20,20 +20,30 @@ input:
         e.g. [ id:'test' ]
   - bam:
       type: file
-      description: Sorted BAM file
+      description: Tumor sample sorted BAM file
       pattern: "*.{bam}"
   - bai:
       type: file
       description: BAM index file
-      pattern: "*.{bai}"
+      pattern: "*.{bam.bai}"
   - intervals:
       type: file
       description: BED file containing target regions for variant calling
       pattern: "*.{bed}"
+  - meta2:
+      type: map
+      description: |
+        Groovy Map containing sample information about the reference fasta
+        e.g. [ id:'reference' ]
   - fasta:
       type: file
       description: Reference genome FASTA file
       pattern: "*.{fasta}"
+  - meta3:
+      type: map
+      description: |
+        Groovy Map containing sample information about the reference fasta fai
+        e.g. [ id:'reference' ]
   - fai:
       type: file
       description: Reference genome FASTA index file
@@ -58,3 +68,5 @@ authors:
 maintainers:
   - "@kaurravneet4123"
   - "@bjohnnyd"
+  - "@nevinwu"
+  - "@AitorPeseta"
diff --git a/modules/nf-core/lofreq/callparallel/tests/main.nf.test b/modules/nf-core/lofreq/callparallel/tests/main.nf.test
@@ -0,0 +1,103 @@
+nextflow_process {
+
+    name "Test Process LOFREQ_CALLPARALLEL"
+    script "../main.nf"
+    process "LOFREQ_CALLPARALLEL"
+
+    tag "modules"
+    tag "modules_nfcore"
+    tag "lofreq"
+    tag "lofreq/callparallel"
+
+    test("sarscov2 - bam") {
+
+        when {
+            process {
+                """
+                input[0] = [ [ id:'test' ], // 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:'fasta' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                        ]
+                input[2] = [ [ id:'fai' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true)
+                        ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.versions).match() },
+                { assert path(process.out.vcf.get(0).get(1)).linesGzip.contains('##INFO=<ID=CONSVAR,Number=0,Type=Flag,Description="Indicates that the variant is a consensus variant (as opposed to a low frequency variant).">') }
+            )
+        }
+
+    }
+
+    test("sarscov2 - bam - bed") {
+
+        when {
+            process {
+                """
+                input[0] = [ [ id:'test' ], // 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),
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/bed/test.bed', checkIfExists: true)
+                        ]
+                input[1] = [ [ id:'fasta' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                        ]
+                input[2] = [ [ id:'fai' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true)
+                        ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out.versions).match() },
+                { assert path(process.out.vcf.get(0).get(1)).linesGzip.contains('##INFO=<ID=CONSVAR,Number=0,Type=Flag,Description="Indicates that the variant is a consensus variant (as opposed to a low frequency variant).">') }
+            )
+        }
+
+    }
+
+    test("sarscov2 - bam - stub") {
+
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [ [ id:'test' ], // 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:'fasta' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                        ]
+                input[2] = [ [ id:'fai' ], // meta map
+                        file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta.fai', checkIfExists: true)
+                        ]
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match() }
+            )
+        }
+
+    }
+
+}
diff --git a/modules/nf-core/lofreq/callparallel/tests/main.nf.test.snap b/modules/nf-core/lofreq/callparallel/tests/main.nf.test.snap
@@ -0,0 +1,59 @@
+{
+    "sarscov2 - bam - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.vcf.gz:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "1": [
+                    "versions.yml:md5,56d45e0015add277b2689f071a4fe3e4"
+                ],
+                "vcf": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.vcf.gz:md5,68b329da9893e34099c7d8ad5cb9c940"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,56d45e0015add277b2689f071a4fe3e4"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "24.04.2"
+        },
+        "timestamp": "2024-07-10T12:57:00.731035"
+    },
+    "sarscov2 - bam": {
+        "content": [
+            [
+                "versions.yml:md5,56d45e0015add277b2689f071a4fe3e4"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "24.04.2"
+        },
+        "timestamp": "2024-07-10T12:59:23.105421"
+    },
+    "sarscov2 - bam - bed": {
+        "content": [
+            [
+                "versions.yml:md5,56d45e0015add277b2689f071a4fe3e4"
+            ]
+        ],
+        "meta": {
+            "nf-test": "0.8.4",
+            "nextflow": "24.04.2"
+        },
+        "timestamp": "2024-07-10T12:59:28.355632"
+    }
+}
diff --git a/modules/nf-core/lofreq/callparallel/tests/tags.yml b/modules/nf-core/lofreq/callparallel/tests/tags.yml
@@ -0,0 +1,2 @@
+lofreq/callparallel:
+  - "modules/nf-core/lofreq/callparallel/**"
diff --git a/modules/nf-core/lofreq/somatic/main.nf b/modules/nf-core/lofreq/somatic/main.nf
@@ -1,14 +1,13 @@
 process LOFREQ_SOMATIC {
     tag "$meta.id"
-    label "process_high"
+    label 'process_high'
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
         'https://depot.galaxyproject.org/singularity/lofreq:2.1.5--py38h588ecb2_4' :
         'biocontainers/lofreq:2.1.5--py38h588ecb2_4' }"
-
     input:
-    tuple val(meta), path(tumor_bam), path(tumor_bai), path(normal_bam), path(normal_bai), path(target_bed)
+    tuple val(meta), path(tumor), path(tumor_index), path(normal), path(normal_index), path(target_bed)
     tuple val(meta2), path(fasta)
     tuple val(meta3), path(fai)
 
@@ -23,17 +22,39 @@ process LOFREQ_SOMATIC {
     def args = task.ext.args ?: ""
     def prefix = task.ext.prefix ?: "${meta.id}"
     def options_target_bed = target_bed ? "-l ${target_bed}" : ""
+
+    def tumor_cram =  tumor.Extension == "cram" ? true : false
+    def normal_cram =  normal.Extension == "cram" ? true : false
+
+    def tumor_out = tumor_cram ? tumor.BaseName + ".bam" : "${tumor}"
+    def normal_out = normal_cram ? normal.BaseName + ".bam" : "${normal}"
+
+    def samtools_cram_convert = ''
+    samtools_cram_convert += normal_cram ? "    samtools view -T $fasta $normal -@ $task.cpus -o $normal_out\n" : ''
+    samtools_cram_convert += normal_cram ? "    samtools index $normal_out\n" : ''
+    samtools_cram_convert += tumor_cram ? "    samtools view -T ${fasta} $tumor -@ $task.cpus -o $tumor_out\n" : ''
+    samtools_cram_convert += tumor_cram ? "    samtools index ${tumor_out}\n" : ''
+
+    def samtools_cram_remove = ''
+    samtools_cram_remove += tumor_cram ? "     rm $tumor_out\n" : ''
+    samtools_cram_remove += tumor_cram ? "     rm ${tumor_out}.bai\n " : ''
+    samtools_cram_remove += normal_cram ? "     rm $normal_out\n" : ''
+    samtools_cram_remove += normal_cram ? "     rm ${normal_out}.bai\n " : ''
     """
+    $samtools_cram_convert
+
     lofreq \\
         somatic \\
         --threads $task.cpus \\
         $args \\
         -f $fasta \\
-        -t $tumor_bam \\
-        -n $normal_bam \\
+        -t $tumor_out \\
+        -n $normal_out \\
         ${options_target_bed} \\
         -o ${prefix}
 
+    $samtools_cram_remove
+
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":
         lofreq: \$(echo \$(lofreq version 2>&1) | sed 's/^version: //; s/ *commit.*\$//')
@@ -43,7 +64,7 @@ process LOFREQ_SOMATIC {
     stub:
     def prefix = task.ext.prefix ?: "${meta.id}"
     """
-    touch ${prefix}.vcf.gz
+    echo "" | gzip > ${prefix}.vcf.gz
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

diff --git a/modules/nf-core/lofreq/somatic/meta.yml b/modules/nf-core/lofreq/somatic/meta.yml
@@ -12,7 +12,7 @@ tools:
       homepage: https://csb5.github.io/lofreq/
       documentation: https://csb5.github.io/lofreq/commands/
       doi: 10.1093/nar/gks918
-      licence: "['MIT']"
+      licence: ["MIT"]
 input:
   - meta:
       type: map
@@ -23,22 +23,22 @@ input:
   - meta3:
       type: map
       description: Groovy Map containing reference information
-  - tumor_bam:
+  - tumor:
       type: file
       description: tumor sample input BAM file
       pattern: "*.{bam}"
-  - tumor_bai:
+  - tumor_index:
       type: file
       description: tumor sample BAM index file
-      pattern: "*.{bai}"
-  - normal_bam:
+      pattern: "*.{bam.bai}"
+  - normal:
       type: file
       description: normal sample input BAM file
       pattern: "*.{bam}"
-  - normal_bai:
+  - normal_index:
       type: file
       description: normal sample BAM index file
-      pattern: "*.{bai}"
+      pattern: "*.{bam.bai}"
   - fasta:
       type: file
       description: Reference genome FASTA file
@@ -69,3 +69,4 @@ authors:
   - "@nevinwu"
 maintainers:
   - "@nevinwu"
+  - "@AitorPeseta"
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,2 @@
		lofreq/callparallel:
		- "modules/nf-core/lofreq/callparallel/**"