Merge branch 'master' into clipkit

nf-core · Nov 14, 2024 · 35dda74 · 35dda74
2 parents 95a278b + 9208952
commit 35dda74
Show file tree

Hide file tree

Showing 49 changed files with 839 additions and 531 deletions.
diff --git a/modules/nf-core/biscuit/align/environment.yml b/modules/nf-core/biscuit/align/environment.yml
@@ -2,5 +2,5 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - bioconda::biscuit=1.1.0.20220707
-  - bioconda::samtools=1.16.1
+  - bioconda::biscuit=1.5.0.20240506
+  - bioconda::samtools=1.21
diff --git a/modules/nf-core/biscuit/align/main.nf b/modules/nf-core/biscuit/align/main.nf
@@ -4,12 +4,13 @@ process BISCUIT_ALIGN {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'https://depot.galaxyproject.org/singularity/mulled-v2-d94f582b04a3edcede1215189c0d881506640fd9:6519548ea4f3d6a526c78ad0350c58f867f28574-0':
-        'biocontainers/mulled-v2-d94f582b04a3edcede1215189c0d881506640fd9:6519548ea4f3d6a526c78ad0350c58f867f28574-0' }"
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/5b/5b542bbe1f99afd494ef07423ea8b52f2b8a081b85f92db2726c283c78da3cf0/data':
+        'community.wave.seqera.io/library/biscuit_samtools:84373c8a97fa63b8' }"
 
     input:
     tuple val(meta), path(reads)
-    path index
+    tuple val(meta2), path(fasta)
+    tuple val(meta3), path(index)
 
     output:
     tuple val(meta), path("*.bam"), emit: bam
@@ -23,18 +24,34 @@ process BISCUIT_ALIGN {
     def args = task.ext.args ?: ''
     def args2 = task.ext.args2 ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
-    def biscuit_cpus = (int) Math.max(Math.floor(task.cpus*0.9),1)
-    def samtools_cpus = task.cpus-biscuit_cpus
+    def biscuit_cpus = [(task.cpus * 0.9) as int, 1].max()
+    def samtools_cpus = (task.cpus - biscuit_cpus < 1) ? biscuit_cpus : (task.cpus - biscuit_cpus)
     """
-    INDEX=`find -L ./ -name "*.bis.amb" | sed 's/\\.bis.amb\$//'`
+    ln -sf \$(readlink $fasta) $index/$fasta
 
     biscuit align \\
         $args \\
         -@ $biscuit_cpus \\
-        \$INDEX \\
+        $index/$fasta \\
         $reads \\
-        | samtools sort $args2 --threads $samtools_cpus --write-index -o ${prefix}.bam##idx##${prefix}.bam.bai -
+        | samtools sort $args2 --threads $samtools_cpus --write-index -o ${prefix}.bam -O BAM -
 
+    samtools index ${prefix}.bam
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        biscuit: \$( biscuit version |& sed '1!d; s/^.*BISCUIT Version: //' )
+        samtools: \$( samtools --version |& sed '1!d; s/^.*samtools //' )
+    END_VERSIONS
+    """
+
+    stub:
+    def args = task.ext.args ?: ''
+    def args2 = task.ext.args2 ?: ''
+    def prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    touch ${prefix}.bam
+    touch ${prefix}.bam.bai
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":

diff --git a/modules/nf-core/biscuit/align/meta.yml b/modules/nf-core/biscuit/align/meta.yml
@@ -27,12 +27,24 @@ input:
     - reads:
         type: file
         description: |
-          List of input fastq files of size 1 and 2 for single-end and paired-end data,
+          List of input FastQ files of size 1 and 2 for single-end and paired-end data,
           respectively.
-  - - index:
+  - - meta2:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. [ id:'test', single_end:false ]
+    - fasta:
+        type: file
+        description: Input genome fasta file
+  - - meta3:
+        type: map
+        description: |
+          Groovy Map containing sample information
+          e.g. [ id:'test', single_end:false ]
+    - index:
         type: directory
-        description: Biscuit genome index directory (generated with 'biscuit index')
-        pattern: "BiscuitIndex"
+        description: Directory containing biscuit genome index
 output:
   - bam:
       - meta:
@@ -63,3 +75,4 @@ authors:
   - "@njspix"
 maintainers:
   - "@njspix"
+  - "@sateeshperi"
diff --git a/modules/nf-core/biscuit/align/tests/main.nf.test b/modules/nf-core/biscuit/align/tests/main.nf.test
@@ -12,16 +12,19 @@ nextflow_process {
 
     setup {
         run("BISCUIT_INDEX") {
-            script "../../index/main.nf"
+            script "../../../biscuit/index/main.nf"
             process {
             """
-            input[0] = file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+            input[0] = Channel.of([
+                    [ id:'test' ], // meta map
+                    file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+            ])
             """
         }
         }
     }
 
-    test("Single-End sarscov2 test_methylated_1 [fastq_gz]") {
+    test("sarscov2 single-end [fastq_gz]") {
 
         when {
             process {
@@ -30,22 +33,29 @@ nextflow_process {
                             [ id:'test' ], // meta map
                             [ file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test.methylated_1.fastq.gz', checkIfExists: true) ]
                 ]
-                input[1] = BISCUIT_INDEX.out.index
+                input[1] = Channel.of([
+                                [ id:'test' ], // meta map
+                                file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                ])
+                input[2] = BISCUIT_INDEX.out.index
                 """
             }
         }
 
         then {
             assertAll(
                 { assert process.success },
-                { assert file(process.out.bam[0][1]).name == "test.bam" },
-                { assert file(process.out.bai[0][1]).name == "test.bam.bai" },
-                { assert snapshot(process.out.versions).match("single-end-versions") }
+                { assert snapshot(
+					bam(process.out.bam[0][1]).getReadsMD5(),
+                    process.out.bai,
+					process.out.versions
+					).match()
+				}
           )
         }
     }
 
-    test("Paired-End sarscov2 test_methylated_1 [fastq_gz]") {
+    test("sarscov2 paired-end [fastq_gz]") {
 
         when {
             process {
@@ -55,17 +65,53 @@ nextflow_process {
                             [ file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test.methylated_1.fastq.gz', checkIfExists: true),
                               file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test.methylated_2.fastq.gz', checkIfExists: true) ]
                 ]
-                input[1] = BISCUIT_INDEX.out.index
+                input[1] = Channel.of([
+                                [ id:'test' ], // meta map
+                                file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                ])
+                input[2] = BISCUIT_INDEX.out.index
                 """
             }
         }
 
         then {
             assertAll(
                 { assert process.success },
-                { assert file(process.out.bam[0][1]).name == "test.bam" },
-                { assert file(process.out.bai[0][1]).name == "test.bam.bai" },
-                { assert snapshot(process.out.versions).match("paired-end-versions") }
+                { assert snapshot(
+					bam(process.out.bam[0][1]).getReadsMD5(),
+                    process.out.bai,
+					process.out.versions
+					).match()
+				}
+          )
+        }
+    }
+
+    test("sarscov2 paired-end [fastq_gz] - stub") {
+        options "-stub"
+
+        when {
+            process {
+                """
+                input[0] = [
+                            [ id:'test' ], // meta map
+                            [ file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test.methylated_1.fastq.gz', checkIfExists: true),
+                              file(params.modules_testdata_base_path + 'genomics/sarscov2/illumina/fastq/test.methylated_2.fastq.gz', checkIfExists: true) ]
+                ]
+                input[1] = Channel.of([
+                                [ id:'test' ], // meta map
+                                file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/genome.fasta', checkIfExists: true)
+                ])
+                input[2] = BISCUIT_INDEX.out.index
+                """
+            }
+        }
+
+        then {
+            assertAll(
+                { assert process.success },
+                { assert snapshot(process.out).match()
+				}
           )
         }
     }

diff --git a/modules/nf-core/biscuit/align/tests/main.nf.test.snap b/modules/nf-core/biscuit/align/tests/main.nf.test.snap
@@ -1,26 +1,93 @@
 {
-    "single-end-versions": {
+    "sarscov2 single-end [fastq_gz]": {
         "content": [
+            "499055250804bb3674c724bd9adf9d02",
             [
-                "versions.yml:md5,0b2a8c4c438a785a890110a458399bfd"
+                [
+                    {
+                        "id": "test"
+                    },
+                    "test.bam.bai:md5,a748afc3cec6c27906ae80638e2024f5"
+                ]
+            ],
+            [
+                "versions.yml:md5,18c276eafa69dc43fa4d5b83b8392016"
             ]
         ],
         "meta": {
-            "nf-test": "0.8.4",
-            "nextflow": "24.04.1"
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.0"
         },
-        "timestamp": "2024-05-22T20:06:02.295706212"
+        "timestamp": "2024-11-14T10:03:51.461022384"
     },
-    "paired-end-versions": {
+    "sarscov2 paired-end [fastq_gz]": {
         "content": [
+            "6b5afc62da5533472e1479950bb5f09a",
+            [
+                [
+                    {
+                        "id": "test"
+                    },
+                    "test.bam.bai:md5,ba60b802f583496e330c555bca9f0b97"
+                ]
+            ],
             [
-                "versions.yml:md5,0b2a8c4c438a785a890110a458399bfd"
+                "versions.yml:md5,18c276eafa69dc43fa4d5b83b8392016"
             ]
         ],
         "meta": {
-            "nf-test": "0.8.4",
-            "nextflow": "24.04.1"
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.0"
+        },
+        "timestamp": "2024-11-14T10:04:02.88266119"
+    },
+    "sarscov2 paired-end [fastq_gz] - stub": {
+        "content": [
+            {
+                "0": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "1": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.bam.bai:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "2": [
+                    "versions.yml:md5,18c276eafa69dc43fa4d5b83b8392016"
+                ],
+                "bai": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.bam.bai:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "bam": [
+                    [
+                        {
+                            "id": "test"
+                        },
+                        "test.bam:md5,d41d8cd98f00b204e9800998ecf8427e"
+                    ]
+                ],
+                "versions": [
+                    "versions.yml:md5,18c276eafa69dc43fa4d5b83b8392016"
+                ]
+            }
+        ],
+        "meta": {
+            "nf-test": "0.9.2",
+            "nextflow": "24.10.0"
         },
-        "timestamp": "2024-05-22T20:06:17.879689395"
+        "timestamp": "2024-11-13T06:48:19.558887499"
     }
 }
diff --git a/modules/nf-core/biscuit/align/tests/tags.yml b/modules/nf-core/biscuit/align/tests/tags.yml
diff --git a/modules/nf-core/biscuit/biscuitblaster/environment.yml b/modules/nf-core/biscuit/biscuitblaster/environment.yml
@@ -2,6 +2,6 @@ channels:
   - conda-forge
   - bioconda
 dependencies:
-  - bioconda::biscuit=1.1.0.20220707
+  - bioconda::biscuit=1.5.0.20240506
   - bioconda::samblaster=0.1.26
-  - bioconda::samtools=1.16.1
+  - bioconda::samtools=1.21
diff --git a/modules/nf-core/biscuit/biscuitblaster/main.nf b/modules/nf-core/biscuit/biscuitblaster/main.nf
@@ -4,12 +4,13 @@ process BISCUIT_BLASTER {
 
     conda "${moduleDir}/environment.yml"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'https://depot.galaxyproject.org/singularity/mulled-v2-d94f582b04a3edcede1215189c0d881506640fd9:6519548ea4f3d6a526c78ad0350c58f867f28574-0':
-        'biocontainers/mulled-v2-d94f582b04a3edcede1215189c0d881506640fd9:6519548ea4f3d6a526c78ad0350c58f867f28574-0' }"
+        'https://community-cr-prod.seqera.io/docker/registry/v2/blobs/sha256/22/22f141a92e12f98040054b9a47a51fb4de9c544174bbfa6d251a110f5e26fddc/data':
+        'community.wave.seqera.io/library/biscuit_samblaster_samtools:a683d0887e7a91bf' }"
 
     input:
     tuple val(meta), path(reads)
-    path index
+    tuple val(meta2), path(fasta)
+    tuple val(meta3), path(index)
 
     output:
     tuple val(meta), path("*.bam"), emit: bam
@@ -24,23 +25,24 @@ process BISCUIT_BLASTER {
     def args = task.ext.args ?: ''
     def args2 = task.ext.args2 ?: ''
     def args3 = task.ext.args3 ?: ''
-    def biscuit_cpus = (int) Math.max(Math.floor(task.cpus*0.95),1)
-    def samtools_cpus = task.cpus-biscuit_cpus
+    def biscuit_cpus = [(task.cpus * 0.9) as int, 1].max()
+    def samtools_cpus = (task.cpus - biscuit_cpus < 1) ? biscuit_cpus : (task.cpus - biscuit_cpus)
     """
-    INDEX=`find -L ./ -name "*.bis.amb" | sed 's/\\.bis.amb\$//'`
+    ln -sf \$(readlink $fasta) $index/$fasta
 
     biscuit align \\
         -@ $biscuit_cpus \\
         $args \\
-        \$INDEX \\
-        $reads | \\
-    samblaster \\
-        $args2 | \\
-    samtools sort \\
-        -@ $samtools_cpus \\
-        $args3 \\
-        --write-index \\
-        -o ${prefix}.bam##idx##${prefix}.bam.bai
+        $index/$fasta\\
+        $reads \\
+        | samblaster $args2 \\
+        | samtools sort \\
+            -@ $samtools_cpus \\
+            $args3 \\
+            --write-index \\
+            -o ${prefix}.bam -O BAM -
+
+    samtools index ${prefix}.bam
 
     cat <<-END_VERSIONS > versions.yml
     "${task.process}":