From 8a56d02943b281b1a8fedbe9928b9b8656b40eb1 Mon Sep 17 00:00:00 2001
From: jsabban <jules.sabban@inrae.fr>
Date: Thu, 27 Apr 2023 15:57:52 +0200
Subject: [PATCH 1/5] Improve fastq detection
Ref : #56
---
workflow/illumina_qc.nf | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/workflow/illumina_qc.nf b/workflow/illumina_qc.nf
index 8bdf8f6..e95cb13 100644
--- a/workflow/illumina_qc.nf
+++ b/workflow/illumina_qc.nf
@@ -33,7 +33,7 @@ ch_DemuxStatXML=Channel.fromPath(params.inputdir+'/Stats/DemultiplexingStats.xml
// fastq one by one
ch_read=Channel
- .fromPath(params.data_location+'/**_R{1,2}_*.fastq.gz')
+ .fromPath(params.data_location+'/**_R{1,2}(_*)?.fastq.gz')
.map{$it -> [$it.simpleName, $it]}
// fastq paired
--
GitLab
From f869d1a9f079d9002e940f549386773c0bdd12ba Mon Sep 17 00:00:00 2001
From: jsabban <jules.sabban@inrae.fr>
Date: Thu, 27 Apr 2023 16:01:30 +0200
Subject: [PATCH 2/5] Memory for sortMeRNA increase with retry
Ref: #54
---
conf/base.config | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/conf/base.config b/conf/base.config
index d330b07..d375699 100644
--- a/conf/base.config
+++ b/conf/base.config
@@ -278,7 +278,7 @@ process {
withName: SORTMERNA {
module = 'bioinfo/sortmerna-4.3.2'
- memory = '1.GB'
+ memory = { 2.GB * task.attempt }
time = { 10.h * task.attempt }
cpus = { 1 * task.attempt }
--
GitLab
From e0ec295559d2a4e3c3457133112975c66fd55097 Mon Sep 17 00:00:00 2001
From: jsabban <jules.sabban@inrae.fr>
Date: Fri, 5 May 2023 15:35:44 +0200
Subject: [PATCH 3/5] Improve glob pattern of fastq
Ref: #56
---
conf/base.config | 7 +++----
workflow/illumina_qc.nf | 2 +-
2 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/conf/base.config b/conf/base.config
index d375699..b99f889 100644
--- a/conf/base.config
+++ b/conf/base.config
@@ -75,7 +75,7 @@ import java.nio.file.Files
import java.nio.file.Paths
def n_read_files = Files.walk(Paths.get(params.data_location))
.filter(Files::isRegularFile)
- .filter(p -> p.getFileName().toString().matches(".*_R[12]_.*\\.fastq\\.gz"))
+ .filter(p -> p.getFileName().toString().matches(".*_R[12](_.*)?\\.fastq\\.gz"))
.count()
params.n_samples = n_read_files / 2
@@ -153,10 +153,9 @@ process {
saveAs: { filename -> "${name}.html" }
]
- errorStrategy = { task.exitStatus in [143,137,104,134,139] ? 'retry' : 'finish' }
- maxRetries = 3
+ maxRetries = 4
module = ['bioinfo/FastQC_v0.11.7']
- time = { 1.h * task.attempt * params.resource_factor }
+ time = { 2.h * task.attempt * params.resource_factor }
}
withName: FASTQSCREEN {
diff --git a/workflow/illumina_qc.nf b/workflow/illumina_qc.nf
index e95cb13..344ef6c 100644
--- a/workflow/illumina_qc.nf
+++ b/workflow/illumina_qc.nf
@@ -33,7 +33,7 @@ ch_DemuxStatXML=Channel.fromPath(params.inputdir+'/Stats/DemultiplexingStats.xml
// fastq one by one
ch_read=Channel
- .fromPath(params.data_location+'/**_R{1,2}(_*)?.fastq.gz')
+ .fromPath(params.data_location+'/**_R{1,2}{_*,*}.fastq.gz')
.map{$it -> [$it.simpleName, $it]}
// fastq paired
--
GitLab
From 8dadf4ad26dc9c7f7569b6d26b4e4ce60f86dffe Mon Sep 17 00:00:00 2001
From: jsabban <jules.sabban@inrae.fr>
Date: Fri, 5 May 2023 15:39:02 +0200
Subject: [PATCH 4/5] Add Fil at readset level
Ref: #53
---
sub-workflows/local/core_pipeline.nf | 1 +
workflow/illumina_qc.nf | 5 +++--
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/sub-workflows/local/core_pipeline.nf b/sub-workflows/local/core_pipeline.nf
index 96b89a8..edbb825 100644
--- a/sub-workflows/local/core_pipeline.nf
+++ b/sub-workflows/local/core_pipeline.nf
@@ -66,4 +66,5 @@ workflow CORE {
fastqscreen_report = FASTQSCREEN.out.report ?: Channel.empty()
fastp_report = DUPLICATED_READS.out.json
subset_fastq = unzip_reads_split.small.mix(SEQTK_SAMPLE.out)
+ fastq_md5 = MD5SUM.out
}
diff --git a/workflow/illumina_qc.nf b/workflow/illumina_qc.nf
index 344ef6c..04c0e66 100644
--- a/workflow/illumina_qc.nf
+++ b/workflow/illumina_qc.nf
@@ -65,7 +65,7 @@ include { RNA_QC } from "$baseDir/sub-workflows/local/rna_qc.nf"
include { MULTIQC } from "${params.shared_modules}/multiqc.nf"
include { workflow_summary as WORKFLOW_SUMMARY } from "${params.shared_modules}/workflow_summary.nf"
include { UPDATE_NGLBI_STATE_FROM_FILE as UPDATE_STATE_FQC } from "${params.shared_modules}/ngl_bi.nf"
-
+include { READSET_FILE_FROM_FILE as ADD_RS_RAW_FILES } from "${params.shared_modules}/ngl_bi.nf" addParams(ext: 'RAW')
// -------------------------------------------------
// EMAIL ON START
// -------------------------------------------------
@@ -90,7 +90,7 @@ workflow ILLUMINA_QC {
} else {
fastq = ch_read
}
-
+
CORE(fastq)
if (params.data_nature == 'DNA') {
@@ -120,6 +120,7 @@ workflow ILLUMINA_QC {
)
if (params.insert_to_ngl){
+ ADD_RS_RAW_FILES(NGLBI.out.readsetsFile, CORE.out.fastq_md5, 'RAW', NGLBI.out.ready)
UPDATE_STATE_FQC(NGLBI.out.readsetsFile, 'F-QC', MULTIQC.out.html)
}
--
GitLab
From 56c13e4378c686672df26dc9405753d9b723be8a Mon Sep 17 00:00:00 2001
From: jsabban <jules.sabban@inrae.fr>
Date: Fri, 5 May 2023 15:42:00 +0200
Subject: [PATCH 5/5] Add .gitignore
---
.gitignore | 2 ++
1 file changed, 2 insertions(+)
create mode 100644 .gitignore
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..6ae31d6
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+dev_DnaPrep.nf
+test.nf
\ No newline at end of file
--
GitLab