From a547b876299b690f3f448fc8cb97ce3aca318dfe Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 08:34:05 +0000
Subject: [PATCH 01/11] version bump
---
.../nf-core/diamond/blastp/environment.yml | 2 +-
modules/nf-core/diamond/blastp/main.nf | 4 ++--
.../diamond/blastp/tests/main.nf.test.snap | 24 +++++++++----------
3 files changed, 15 insertions(+), 15 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/environment.yml b/modules/nf-core/diamond/blastp/environment.yml
index 950c3c5c55e..f19483fe91b 100644
--- a/modules/nf-core/diamond/blastp/environment.yml
+++ b/modules/nf-core/diamond/blastp/environment.yml
@@ -2,4 +2,4 @@ channels:
- conda-forge
- bioconda
dependencies:
- - bioconda::diamond=2.1.8
+ - bioconda::diamond=2.1.11
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index dc01cdcc08c..a43e714ba52 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -4,8 +4,8 @@ process DIAMOND_BLASTP {
conda "${moduleDir}/environment.yml"
container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
- 'https://depot.galaxyproject.org/singularity/diamond:2.1.8--h43eeafb_0' :
- 'biocontainers/diamond:2.1.8--h43eeafb_0' }"
+ 'https://depot.galaxyproject.org/singularity/diamond:2.1.11--h5ca1c30_0' :
+ 'biocontainers/diamond:2.1.11--h5ca1c30_0' }"
input:
tuple val(meta) , path(fasta)
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
index e323c8b89c1..47c4758f66d 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
@@ -10,14 +10,14 @@
]
],
[
- "versions.yml:md5,57a0ebeb0a8a732c941ae0102639a9d0"
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
]
],
"meta": {
- "nf-test": "0.8.4",
- "nextflow": "24.04.3"
+ "nf-test": "0.9.2",
+ "nextflow": "24.10.2"
},
- "timestamp": "2024-07-29T14:40:23.906848"
+ "timestamp": "2025-01-28T08:32:39.553875424"
},
"gz_txt": {
"content": [
@@ -30,25 +30,25 @@
]
],
[
- "versions.yml:md5,57a0ebeb0a8a732c941ae0102639a9d0"
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
]
],
"meta": {
- "nf-test": "0.8.4",
- "nextflow": "24.04.3"
+ "nf-test": "0.9.2",
+ "nextflow": "24.10.2"
},
- "timestamp": "2024-07-29T14:40:29.865487"
+ "timestamp": "2025-01-28T08:32:49.763438642"
},
"Should search for protein hits against a DIAMOND db and return a daa format file of hits": {
"content": [
[
- "versions.yml:md5,57a0ebeb0a8a732c941ae0102639a9d0"
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
]
],
"meta": {
- "nf-test": "0.8.4",
- "nextflow": "24.04.3"
+ "nf-test": "0.9.2",
+ "nextflow": "24.10.2"
},
- "timestamp": "2024-07-29T14:40:35.362027"
+ "timestamp": "2025-01-28T08:32:58.984888127"
}
}
\ No newline at end of file
From 06d1bfa64f3f4800e44222d47f8b2e794d945c91 Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 08:39:46 +0000
Subject: [PATCH 02/11] remove unnecessary unzipping of gz fasta files
---
modules/nf-core/diamond/blastp/main.nf | 8 +-------
1 file changed, 1 insertion(+), 7 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index a43e714ba52..2fa412a9369 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -29,8 +29,6 @@ process DIAMOND_BLASTP {
script:
def args = task.ext.args ?: ''
def prefix = task.ext.prefix ?: "${meta.id}"
- def is_compressed = fasta.getExtension() == "gz" ? true : false
- def fasta_name = is_compressed ? fasta.getBaseName() : fasta
def columns = blast_columns ? "${blast_columns}" : ''
switch ( out_ext ) {
case "blast": outfmt = 0; break
@@ -47,17 +45,13 @@ process DIAMOND_BLASTP {
break
}
"""
- if [ "${is_compressed}" == "true" ]; then
- gzip -c -d ${fasta} > ${fasta_name}
- fi
-
DB=`find -L ./ -name "*.dmnd" | sed 's/\\.dmnd\$//'`
diamond \\
blastp \\
--threads ${task.cpus} \\
--db \$DB \\
- --query ${fasta_name} \\
+ --query ${fasta} \\
--outfmt ${outfmt} ${columns} \\
${args} \\
--out ${prefix}.${out_ext}
From 7e3563b085faab8f9031494220dda7412cd7048c Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 08:41:19 +0000
Subject: [PATCH 03/11] label updated to process_high
---
modules/nf-core/diamond/blastp/main.nf | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index 2fa412a9369..e68602afcc0 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -1,6 +1,6 @@
process DIAMOND_BLASTP {
tag "$meta.id"
- label 'process_medium'
+ label 'process_high'
conda "${moduleDir}/environment.yml"
container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
From d7751f61637cfcd67d3839f2a8323484de0859bb Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 08:50:19 +0000
Subject: [PATCH 04/11] db from arg and switch changed to if else
---
modules/nf-core/diamond/blastp/main.nf | 35 ++++++++++++++------------
1 file changed, 19 insertions(+), 16 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index e68602afcc0..5677f6f34b4 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -30,27 +30,30 @@ process DIAMOND_BLASTP {
def args = task.ext.args ?: ''
def prefix = task.ext.prefix ?: "${meta.id}"
def columns = blast_columns ? "${blast_columns}" : ''
- switch ( out_ext ) {
- case "blast": outfmt = 0; break
- case "xml": outfmt = 5; break
- case "txt": outfmt = 6; break
- case "daa": outfmt = 100; break
- case "sam": outfmt = 101; break
- case "tsv": outfmt = 102; break
- case "paf": outfmt = 103; break
- default:
- outfmt = '6';
- out_ext = 'txt';
- log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)");
- break
+ if (out_ext == "blast") {
+ outfmt = 0
+ } else if (out_ext == "xml") {
+ outfmt = 5
+ } else if (out_ext == "txt") {
+ outfmt = 6
+ } else if (out_ext == "daa") {
+ outfmt = 100
+ } else if (out_ext == "sam") {
+ outfmt = 101
+ } else if (out_ext == "tsv") {
+ outfmt = 102
+ } else if (out_ext == "paf") {
+ outfmt = 103
+ } else {
+ outfmt = '6'
+ out_ext = 'txt'
+ log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
"""
- DB=`find -L ./ -name "*.dmnd" | sed 's/\\.dmnd\$//'`
-
diamond \\
blastp \\
--threads ${task.cpus} \\
- --db \$DB \\
+ --db ${db} \\
--query ${fasta} \\
--outfmt ${outfmt} ${columns} \\
${args} \\
From bfbd61a4e62ea925cf0b1e282782edb129f057a5 Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 09:18:22 +0000
Subject: [PATCH 05/11] stub updated with if else
---
modules/nf-core/diamond/blastp/main.nf | 32 +++++++++++++++-----------
1 file changed, 18 insertions(+), 14 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index 5677f6f34b4..df418308e6e 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -66,21 +66,25 @@ process DIAMOND_BLASTP {
"""
stub:
- def args = task.ext.args ?: ''
def prefix = task.ext.prefix ?: "${meta.id}"
- switch ( out_ext ) {
- case "blast": outfmt = 0; break
- case "xml": outfmt = 5; break
- case "txt": outfmt = 6; break
- case "daa": outfmt = 100; break
- case "sam": outfmt = 101; break
- case "tsv": outfmt = 102; break
- case "paf": outfmt = 103; break
- default:
- outfmt = '6';
- out_ext = 'txt';
- log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)");
- break
+ if (out_ext == "blast") {
+ outfmt = 0
+ } else if (out_ext == "xml") {
+ outfmt = 5
+ } else if (out_ext == "txt") {
+ outfmt = 6
+ } else if (out_ext == "daa") {
+ outfmt = 100
+ } else if (out_ext == "sam") {
+ outfmt = 101
+ } else if (out_ext == "tsv") {
+ outfmt = 102
+ } else if (out_ext == "paf") {
+ outfmt = 103
+ } else {
+ outfmt = '6'
+ out_ext = 'txt'
+ log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
"""
From 62fc882feb693ac92dd159927b192952d772e67b Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 10:28:48 +0000
Subject: [PATCH 06/11] allowing gz outputs, test names updated and added stub
test
---
modules/nf-core/diamond/blastp/main.nf | 18 +-
.../nf-core/diamond/blastp/tests/main.nf.test | 47 ++--
.../diamond/blastp/tests/main.nf.test.snap | 217 ++++++++++++++++--
3 files changed, 232 insertions(+), 50 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index df418308e6e..a1480358b86 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -14,14 +14,14 @@ process DIAMOND_BLASTP {
val blast_columns
output:
- tuple val(meta), path('*.blast'), optional: true, emit: blast
- tuple val(meta), path('*.xml') , optional: true, emit: xml
- tuple val(meta), path('*.txt') , optional: true, emit: txt
- tuple val(meta), path('*.daa') , optional: true, emit: daa
- tuple val(meta), path('*.sam') , optional: true, emit: sam
- tuple val(meta), path('*.tsv') , optional: true, emit: tsv
- tuple val(meta), path('*.paf') , optional: true, emit: paf
- path "versions.yml" , emit: versions
+ tuple val(meta), path('*.{blast,blast.gz}'), optional: true, emit: blast
+ tuple val(meta), path('*.{xml,xml.gz}') , optional: true, emit: xml
+ tuple val(meta), path('*.{txt,txt.gz}') , optional: true, emit: txt
+ tuple val(meta), path('*.{daa,daa.gz}') , optional: true, emit: daa
+ tuple val(meta), path('*.{sam,sam.gz}') , optional: true, emit: sam
+ tuple val(meta), path('*.{tsv,tsv.gz}') , optional: true, emit: tsv
+ tuple val(meta), path('*.{paf,paf.gz}') , optional: true, emit: paf
+ path "versions.yml" , emit: versions
when:
task.ext.when == null || task.ext.when
@@ -49,6 +49,8 @@ process DIAMOND_BLASTP {
out_ext = 'txt'
log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
+ if ( args =~ /--compress\s+1/ ) out_ext += '.gz'
+
"""
diamond \\
blastp \\
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test b/modules/nf-core/diamond/blastp/tests/main.nf.test
index f21e926de9d..c7c1deba330 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test
@@ -23,7 +23,7 @@ nextflow_process {
}
}
- test("Should search for protein hits against a DIAMOND db and return a tab separated output file of hits") {
+ test("sarscov2 - proteome - txt") {
when {
params {
@@ -42,17 +42,13 @@ nextflow_process {
then {
assertAll(
{ assert process.success },
- { assert snapshot(
- process.out.txt,
- process.out.versions
- ).match()
- }
+ { assert snapshot(process.out).match() }
)
}
}
- test("Should search for zipped protein hits against a DIAMOND db and return a tab separated output file of hits") {
+ test("sarscov2 - proteome - gz - txt") {
when {
params {
@@ -71,22 +67,15 @@ nextflow_process {
then {
assertAll(
{ assert process.success },
- { assert snapshot(
- process.out.txt,
- process.out.versions
- ).match("gz_txt")
- }
+ { assert snapshot(process.out).match("gz_txt")}
)
}
}
- test("Should search for protein hits against a DIAMOND db and return a daa format file of hits") {
+ test("sarscov2 - proteome - daa") {
when {
- params {
- outdir = "$outputDir"
- }
process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
@@ -101,7 +90,31 @@ nextflow_process {
assertAll(
{ assert process.success },
{ assert process.out.daa },
- { assert snapshot(process.out.versions).match() }
+ { assert snapshot(process.out.versions).match("daa") }
+ )
+ }
+
+ }
+
+ test("sarscov2 - proteome - stub") {
+
+ options "-stub"
+
+ when {
+ process {
+ """
+ input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
+ input[1] = DIAMOND_MAKEDB.out.db
+ input[2] = 'txt'
+ input[3] = 'qseqid qlen'
+ """
+ }
+ }
+
+ then {
+ assertAll(
+ { assert process.success },
+ { assert snapshot(process.out).match("stub") }
)
}
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
index 47c4758f66d..b34ef399022 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
@@ -1,45 +1,143 @@
{
- "Should search for protein hits against a DIAMOND db and return a tab separated output file of hits": {
+ "sarscov2 - proteome - txt": {
"content": [
- [
- [
- {
- "id": "test"
- },
- "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ {
+ "0": [
+
+ ],
+ "1": [
+
+ ],
+ "2": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "3": [
+
+ ],
+ "4": [
+
+ ],
+ "5": [
+
+ ],
+ "6": [
+
+ ],
+ "7": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "blast": [
+
+ ],
+ "daa": [
+
+ ],
+ "paf": [
+
+ ],
+ "sam": [
+
+ ],
+ "tsv": [
+
+ ],
+ "txt": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "versions": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "xml": [
+
]
- ],
- [
- "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
- ]
+ }
],
"meta": {
"nf-test": "0.9.2",
"nextflow": "24.10.2"
},
- "timestamp": "2025-01-28T08:32:39.553875424"
+ "timestamp": "2025-01-28T10:25:13.48912978"
},
"gz_txt": {
"content": [
- [
- [
- {
- "id": "test"
- },
- "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ {
+ "0": [
+
+ ],
+ "1": [
+
+ ],
+ "2": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "3": [
+
+ ],
+ "4": [
+
+ ],
+ "5": [
+
+ ],
+ "6": [
+
+ ],
+ "7": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "blast": [
+
+ ],
+ "daa": [
+
+ ],
+ "paf": [
+
+ ],
+ "sam": [
+
+ ],
+ "tsv": [
+
+ ],
+ "txt": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "versions": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "xml": [
+
]
- ],
- [
- "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
- ]
+ }
],
"meta": {
"nf-test": "0.9.2",
"nextflow": "24.10.2"
},
- "timestamp": "2025-01-28T08:32:49.763438642"
+ "timestamp": "2025-01-28T10:25:20.993203497"
},
- "Should search for protein hits against a DIAMOND db and return a daa format file of hits": {
+ "daa": {
"content": [
[
"versions.yml:md5,5f638327037bee3c00e17521c04a652f"
@@ -49,6 +147,75 @@
"nf-test": "0.9.2",
"nextflow": "24.10.2"
},
- "timestamp": "2025-01-28T08:32:58.984888127"
+ "timestamp": "2025-01-28T10:25:28.126992812"
+ },
+ "stub": {
+ "content": [
+ {
+ "0": [
+
+ ],
+ "1": [
+
+ ],
+ "2": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,d41d8cd98f00b204e9800998ecf8427e"
+ ]
+ ],
+ "3": [
+
+ ],
+ "4": [
+
+ ],
+ "5": [
+
+ ],
+ "6": [
+
+ ],
+ "7": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "blast": [
+
+ ],
+ "daa": [
+
+ ],
+ "paf": [
+
+ ],
+ "sam": [
+
+ ],
+ "tsv": [
+
+ ],
+ "txt": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt:md5,d41d8cd98f00b204e9800998ecf8427e"
+ ]
+ ],
+ "versions": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "xml": [
+
+ ]
+ }
+ ],
+ "meta": {
+ "nf-test": "0.9.2",
+ "nextflow": "24.10.2"
+ },
+ "timestamp": "2025-01-28T10:25:34.911633513"
}
}
\ No newline at end of file
From d756499357918cf86f75f825813fba2b6002b05f Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 10:37:36 +0000
Subject: [PATCH 07/11] test for --compress flag added
---
.../nf-core/diamond/blastp/tests/main.nf.test | 30 ++++++--
.../diamond/blastp/tests/main.nf.test.snap | 69 +++++++++++++++++++
.../diamond/blastp/tests/nextflow.config | 7 ++
3 files changed, 100 insertions(+), 6 deletions(-)
create mode 100644 modules/nf-core/diamond/blastp/tests/nextflow.config
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test b/modules/nf-core/diamond/blastp/tests/main.nf.test
index c7c1deba330..7934334cc55 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test
@@ -26,9 +26,6 @@ nextflow_process {
test("sarscov2 - proteome - txt") {
when {
- params {
- outdir = "$outputDir"
- }
process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
@@ -51,9 +48,6 @@ nextflow_process {
test("sarscov2 - proteome - gz - txt") {
when {
- params {
- outdir = "$outputDir"
- }
process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta.gz', checkIfExists: true) ]
@@ -96,6 +90,30 @@ nextflow_process {
}
+ test("sarscov2 - proteome - txt - gz") {
+
+ config "./nextflow.config"
+
+ when {
+ process {
+ """
+ input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
+ input[1] = DIAMOND_MAKEDB.out.db
+ input[2] = 'txt'
+ input[3] = 'qseqid qlen'
+ """
+ }
+ }
+
+ then {
+ assertAll(
+ { assert process.success },
+ { assert snapshot(process.out).match("txt_gz") }
+ )
+ }
+
+ }
+
test("sarscov2 - proteome - stub") {
options "-stub"
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
index b34ef399022..44d504337b1 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test.snap
@@ -68,6 +68,75 @@
},
"timestamp": "2025-01-28T10:25:13.48912978"
},
+ "txt_gz": {
+ "content": [
+ {
+ "0": [
+
+ ],
+ "1": [
+
+ ],
+ "2": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt.gz:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "3": [
+
+ ],
+ "4": [
+
+ ],
+ "5": [
+
+ ],
+ "6": [
+
+ ],
+ "7": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "blast": [
+
+ ],
+ "daa": [
+
+ ],
+ "paf": [
+
+ ],
+ "sam": [
+
+ ],
+ "tsv": [
+
+ ],
+ "txt": [
+ [
+ {
+ "id": "test"
+ },
+ "test.txt.gz:md5,8131b1afd717f3d5f2f2417c5b562e6e"
+ ]
+ ],
+ "versions": [
+ "versions.yml:md5,5f638327037bee3c00e17521c04a652f"
+ ],
+ "xml": [
+
+ ]
+ }
+ ],
+ "meta": {
+ "nf-test": "0.9.2",
+ "nextflow": "24.10.2"
+ },
+ "timestamp": "2025-01-28T10:36:04.361504205"
+ },
"gz_txt": {
"content": [
{
diff --git a/modules/nf-core/diamond/blastp/tests/nextflow.config b/modules/nf-core/diamond/blastp/tests/nextflow.config
new file mode 100644
index 00000000000..bd28cb1d433
--- /dev/null
+++ b/modules/nf-core/diamond/blastp/tests/nextflow.config
@@ -0,0 +1,7 @@
+process {
+
+ withName: DIAMOND_BLASTP {
+ ext.args = '--compress 1'
+ }
+
+}
From 01bc4b0d15c8050bb0cc7684244354400b203b3c Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Tue, 28 Jan 2025 10:50:27 +0000
Subject: [PATCH 08/11] meta.yml updated to include new gz outputs
---
modules/nf-core/diamond/blastp/meta.yml | 49 ++++++++++++++-----------
1 file changed, 28 insertions(+), 21 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/meta.yml b/modules/nf-core/diamond/blastp/meta.yml
index fbddfbd00f1..295dbf30fdc 100644
--- a/modules/nf-core/diamond/blastp/meta.yml
+++ b/modules/nf-core/diamond/blastp/meta.yml
@@ -54,70 +54,77 @@ output:
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.blast":
- type: file
+ pattern: "*.{blast,blast.gz}"
+ - "*.{blast,blast.gz}":
+ type: map
description: File containing blastp hits
- pattern: "*.{blast}"
+ pattern: "*.{blast,blast.gz}"
- xml:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.xml":
- type: file
+ pattern: "*.{xml,xml.gz}"
+ - "*.{xml,xml.gz}":
+ type: map
description: File containing blastp hits
- pattern: "*.{xml}"
+ pattern: "*.{xml,xml.gz}"
- txt:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.txt":
- type: file
+ pattern: "*.{txt,txt.gz}"
+ - "*.{txt,txt.gz}":
+ type: map
description: File containing hits in tabular BLAST format.
- pattern: "*.{txt}"
+ pattern: "*.{txt,txt.gz}"
- daa:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.daa":
- type: file
+ pattern: "*.{daa,daa.gz}"
+ - "*.{daa,daa.gz}":
+ type: map
description: File containing hits DAA format
- pattern: "*.{daa}"
+ pattern: "*.{daa,daa.gz}"
- sam:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.sam":
- type: file
+ pattern: "*.{sam,sam.gz}"
+ - "*.{sam,sam.gz}":
+ type: map
description: File containing aligned reads in SAM format
- pattern: "*.{sam}"
+ pattern: "*.{sam,sam.gz}"
- tsv:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.tsv":
- type: file
+ pattern: "*.{tsv,tsv.gz}"
+ - "*.{tsv,tsv.gz}":
+ type: map
description: Tab separated file containing taxonomic classification of hits
- pattern: "*.{tsv}"
+ pattern: "*.{tsv,tsv.gz}"
- paf:
- meta:
type: map
description: |
Groovy Map containing sample information
e.g. [ id:'test', single_end:false ]
- - "*.paf":
- type: file
+ pattern: "*.{paf,paf.gz}"
+ - "*.{paf,paf.gz}":
+ type: map
description: File containing aligned reads in pairwise mapping format format
- pattern: "*.{paf}"
+ pattern: "*.{paf,paf.gz}"
- versions:
- versions.yml:
type: file
From 06af3739b7c72e73f2e46f941ec86a825768b28b Mon Sep 17 00:00:00 2001
From: vagkaratzas <vagelaros.gee@gmail.com>
Date: Thu, 30 Jan 2025 13:18:50 +0000
Subject: [PATCH 09/11] outfmt instead of out_ext input arg, meta updated
---
modules/nf-core/diamond/blastp/main.nf | 67 +++++++++++--------
modules/nf-core/diamond/blastp/meta.yml | 21 +++---
.../nf-core/diamond/blastp/tests/main.nf.test | 10 +--
3 files changed, 56 insertions(+), 42 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index a1480358b86..dc712b0bed4 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -10,7 +10,7 @@ process DIAMOND_BLASTP {
input:
tuple val(meta) , path(fasta)
tuple val(meta2), path(db)
- val out_ext
+ val outfmt
val blast_columns
output:
@@ -29,26 +29,30 @@ process DIAMOND_BLASTP {
script:
def args = task.ext.args ?: ''
def prefix = task.ext.prefix ?: "${meta.id}"
+
def columns = blast_columns ? "${blast_columns}" : ''
- if (out_ext == "blast") {
- outfmt = 0
- } else if (out_ext == "xml") {
- outfmt = 5
- } else if (out_ext == "txt") {
- outfmt = 6
- } else if (out_ext == "daa") {
- outfmt = 100
- } else if (out_ext == "sam") {
- outfmt = 101
- } else if (out_ext == "tsv") {
- outfmt = 102
- } else if (out_ext == "paf") {
- outfmt = 103
+ def out_ext = ""
+
+ if (outfmt == 0) {
+ out_ext = "blast"
+ } else if (outfmt == 5) {
+ out_ext = "xml"
+ } else if (outfmt == 6) {
+ out_ext = "txt"
+ } else if (outfmt == 100) {
+ out_ext = "daa"
+ } else if (outfmt == 101) {
+ out_ext = "sam"
+ } else if (outfmt == 102) {
+ out_ext = "tsv"
+ } else if (outfmt == 103) {
+ out_ext = "paf"
} else {
outfmt = '6'
out_ext = 'txt'
log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
+
if ( args =~ /--compress\s+1/ ) out_ext += '.gz'
"""
@@ -69,26 +73,31 @@ process DIAMOND_BLASTP {
stub:
def prefix = task.ext.prefix ?: "${meta.id}"
- if (out_ext == "blast") {
- outfmt = 0
- } else if (out_ext == "xml") {
- outfmt = 5
- } else if (out_ext == "txt") {
- outfmt = 6
- } else if (out_ext == "daa") {
- outfmt = 100
- } else if (out_ext == "sam") {
- outfmt = 101
- } else if (out_ext == "tsv") {
- outfmt = 102
- } else if (out_ext == "paf") {
- outfmt = 103
+
+ def out_ext = ""
+
+ if (outfmt == 0) {
+ out_ext = "blast"
+ } else if (outfmt == 5) {
+ out_ext = "xml"
+ } else if (outfmt == 6) {
+ out_ext = "txt"
+ } else if (outfmt == 100) {
+ out_ext = "daa"
+ } else if (outfmt == 101) {
+ out_ext = "sam"
+ } else if (outfmt == 102) {
+ out_ext = "tsv"
+ } else if (outfmt == 103) {
+ out_ext = "paf"
} else {
outfmt = '6'
out_ext = 'txt'
log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
+ if ( args =~ /--compress\s+1/ ) out_ext += '.gz'
+
"""
touch ${prefix}.${out_ext}
diff --git a/modules/nf-core/diamond/blastp/meta.yml b/modules/nf-core/diamond/blastp/meta.yml
index 295dbf30fdc..6518aa9aedd 100644
--- a/modules/nf-core/diamond/blastp/meta.yml
+++ b/modules/nf-core/diamond/blastp/meta.yml
@@ -33,19 +33,24 @@ input:
type: file
description: File of the indexed DIAMOND database
pattern: "*.dmnd"
- - - out_ext:
- type: string
+ - - outfmt:
+ type: integer
description: |
- Specify the type of output file to be generated. `blast` corresponds to
- BLAST pairwise format. `xml` corresponds to BLAST xml format.
- `txt` corresponds to to BLAST tabular format. `tsv` corresponds to
- taxonomic classification format.
- pattern: "blast|xml|txt|daa|sam|tsv|paf"
+ Specify the type of output file to be generated.
+ 0, .blast, BLAST pairwise format.
+ 5, .xml, BLAST XML format.
+ 6, .txt, BLAST tabular format (default). This format can be customized, the 6 may be followed by a space-separated list of the blast_columns keywords, each specifying a field of the output.
+ 100, .daa, DIAMOND alignment archive (DAA). The DAA format is a proprietary binary format that can subsequently be used to generate other output formats using the view command. It is also supported by MEGAN and allows a quick import of results.
+ 101, .sam, SAM format.
+ 102, .tsv, Taxonomic classification. This format will not print alignments but only a taxonomic classification for each query using the LCA algorithm.
+ 103, .paf, PAF format. The custom fields in the format are AS (bit score), ZR (raw score) and ZE (e-value).
+ pattern: "0|5|6|100|101|102|103"
- - blast_columns:
type: string
description: |
Optional space separated list of DIAMOND tabular BLAST output keywords
- used for in conjunction with the 'txt' out_ext option (--outfmt 6). Options:
+ used in conjunction with the --outfmt 6 option (txt).
+ Options:
qseqid sseqid pident length mismatch gapopen qstart qend sstart send evalue bitscore
output:
- blast:
diff --git a/modules/nf-core/diamond/blastp/tests/main.nf.test b/modules/nf-core/diamond/blastp/tests/main.nf.test
index 7934334cc55..9211915173c 100644
--- a/modules/nf-core/diamond/blastp/tests/main.nf.test
+++ b/modules/nf-core/diamond/blastp/tests/main.nf.test
@@ -30,7 +30,7 @@ nextflow_process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
input[1] = DIAMOND_MAKEDB.out.db
- input[2] = 'txt'
+ input[2] = 6
input[3] = 'qseqid qlen'
"""
}
@@ -52,7 +52,7 @@ nextflow_process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta.gz', checkIfExists: true) ]
input[1] = DIAMOND_MAKEDB.out.db
- input[2] = 'txt'
+ input[2] = 6
input[3] = 'qseqid qlen'
"""
}
@@ -74,7 +74,7 @@ nextflow_process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
input[1] = DIAMOND_MAKEDB.out.db
- input[2] = 'daa'
+ input[2] = 100
input[3] = []
"""
}
@@ -99,7 +99,7 @@ nextflow_process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
input[1] = DIAMOND_MAKEDB.out.db
- input[2] = 'txt'
+ input[2] = 6
input[3] = 'qseqid qlen'
"""
}
@@ -123,7 +123,7 @@ nextflow_process {
"""
input[0] = [ [id:'test'], file(params.modules_testdata_base_path + 'genomics/sarscov2/genome/proteome.fasta', checkIfExists: true) ]
input[1] = DIAMOND_MAKEDB.out.db
- input[2] = 'txt'
+ input[2] = 6
input[3] = 'qseqid qlen'
"""
}
From 68d5789d17289c1bf77edaea38a74813917c7bb9 Mon Sep 17 00:00:00 2001
From: Daniel Lundin <erik.rikard.daniel@gmail.com>
Date: Thu, 30 Jan 2025 14:41:55 +0100
Subject: [PATCH 10/11] Update modules/nf-core/diamond/blastp/main.nf
---
modules/nf-core/diamond/blastp/main.nf | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index dc712b0bed4..1342ee23554 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -48,9 +48,9 @@ process DIAMOND_BLASTP {
} else if (outfmt == 103) {
out_ext = "paf"
} else {
- outfmt = '6'
+ log.warn("Unknown output file format provided (${outfmt}): selecting DIAMOND default of tabular BLAST output (txt)")
+ outfmt = 6
out_ext = 'txt'
- log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
if ( args =~ /--compress\s+1/ ) out_ext += '.gz'
From 3c6bd345b942806d1c74773a756a194efefd7df6 Mon Sep 17 00:00:00 2001
From: Daniel Lundin <erik.rikard.daniel@gmail.com>
Date: Thu, 30 Jan 2025 14:42:04 +0100
Subject: [PATCH 11/11] Update modules/nf-core/diamond/blastp/main.nf
---
modules/nf-core/diamond/blastp/main.nf | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/modules/nf-core/diamond/blastp/main.nf b/modules/nf-core/diamond/blastp/main.nf
index 1342ee23554..6dd8d3925ad 100644
--- a/modules/nf-core/diamond/blastp/main.nf
+++ b/modules/nf-core/diamond/blastp/main.nf
@@ -91,9 +91,9 @@ process DIAMOND_BLASTP {
} else if (outfmt == 103) {
out_ext = "paf"
} else {
- outfmt = '6'
+ log.warn("Unknown output file format provided (${outfmt}): selecting DIAMOND default of tabular BLAST output (txt)")
+ outfmt = 6
out_ext = 'txt'
- log.warn("Unknown output file format provided (${out_ext}): selecting DIAMOND default of tabular BLAST output (txt)")
}
if ( args =~ /--compress\s+1/ ) out_ext += '.gz'