test nextflow dsl2.

README.md

My first workflow using nextflow DSL2.

The workflow takes as input a CSV file with the columns sample,path-to-fq1,path-to-fq2:

• Fastqc is invoked for each fastq

• trimmomatic is invoked for each pair of fastq

• Fastqc is then invoked for each trimmed fastq

download.nf

nextflow.preview.dsl=2

process download {
	tag "download ${url}"
	input:
		tuple url,saveas
	output:
		path("${saveas}")
	script:
	"""
	wget -O "${saveas}" "${url}"
	"""
	}

fastqc.nf

nextflow.preview.dsl=2
params.fastqcadapters="VOID"
params.fastqccontaminants="VOID"

fastqcadapters = file(params.fastqcadapters)
fastqccontaminants = file(params.fastqccontaminants)


process fastqc {
    tag "${prefix} ${sample} ${fq.name}"
    input:
	path contaminants 
        tuple prefix,sample, fq
    output:
	tuple prefix,sample,path("*.zip")
	tuple prefix,sample,fq
    script:
      """
      module load fastqc
      mkdir TMP

	fastqc --dir TMP -o . --noextract --contaminants "${contaminants}" --quiet -f "fastq" ${fq}
      """
}

main.nf

nextflow.preview.dsl=2
params.fastqs=null

include {fastqc as fastqc1} from './fastqc.nf'
include {fastqc as fastqc2} from './fastqc.nf'
include {https://gist.github.com/lindenb/df980f2c660f48735ca0ea7c1a816651} from './trimmomatic.nf'
include {download as wget1} from './download.nf'


workflow {
    contaminants = wget1(["https://raw.githubusercontent.com/csf-ngs/fastqc/master/Contaminants/contaminant_list.txt","contaminants"])

    fastqIn = Channel.fromPath(params.fastqs).
      splitCsv(header: false,sep:',',strip:true).
      map{T->[T[0],file(T[1]),file(T[2])]}.
      flatMap{T->[["before",T[0],T[1]],["before",T[0],T[2]]]}

    trimmed = fastqc1(contaminants,fastqIn)[1] |\
      map {T->[T[1],T[2]]} |\
      groupTuple |\
      map{T->[T[0],T[1][0],T[1][1] ]} |\
      trimmomatic |\
      flatMap{T->[["after",T[0],T[1]],["after",T[0],T[2]]]}
      fastqc2(contaminants,trimmed)
}

Makefile

PREFIX=20201006.DSL2
OUTDIR=work

all: main.nf $(OUTDIR)/jeter.fastqs.list 
	mkdir -p $(OUTDIR)
	nextflow run -with-dag workflow.dot -with-report -with-timeline -with-trace -resume \
		-work-dir "$(OUTDIR)" \
		--fastqs $(OUTDIR)/jeter.fastqs.list \
		$<
	@echo "output is $(OUTDIR)"

$(OUTDIR)/jeter.fastqs.list:
	mkdir -p $(OUTDIR)
	echo "S1,/sandbox/users/lindenbaum-p/src/jvarkit-git/src/test/resources/S1.R1.fq.gz,/sandbox/users/lindenbaum-p/src/jvarkit-git/src/test/resources/S1.R2.fq.gz" > $@
	echo "S2,/sandbox/users/lindenbaum-p/src/jvarkit-git/src/test/resources/S2.R1.fq.gz,/sandbox/users/lindenbaum-p/src/jvarkit-git/src/test/resources/S2.R2.fq.gz" >> $@
	test -s $@

workdir:
	@echo $(OUTDIR)
clean:
	rm -rf "$(OUTDIR)"
springclean:
	rm -rf ".nextflow" report.htm* timeline* trace.tx*   workflow.dot*

trimmomatic.nf

nextflow.preview.dsl=2

process trimmomatic {
    cache "lenient"
    tag "${sample}"
    input:
        tuple sample, fq1, fq2
    output:
	tuple sample,path("${sample}.trimmed1.fq.gz"),path("${sample}.trimmed2.fq.gz")
    script:
      """
      module load trimmomatic

# trim fastq using  trimmomatic

cat << EOF > TruSeq2-PE.fa
>PrefixPE/1
AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT
>PrefixPE/2
CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT
>PCR_Primer1
AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT
>PCR_Primer1_rc
AGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTAGATCTCGGTGGTCGCCGTATCATT
>PCR_Primer2
CAAGCAGAAGACGGCATACGAGATCGGTCTCGGCATTCCTGCTGAACCGCTCTTCCGATCT
>PCR_Primer2_rc
AGATCGGAAGAGCGGTTCAGCAGGAATGCCGAGACCGATCTCGTATGCCGTCTTCTGCTTG
>FlowCell1
TTTTTTTTTTAATGATACGGCGACCACCGAGATCTACAC
>FlowCell2
TTTTTTTTTTCAAGCAGAAGACGGCATACGA
EOF


      java -jar \${TRIMMOMATIC_JAR} PE -phred33 \
	"${fq1}" "${fq2}" \
	"${sample}.trimmed1.fq.gz" "output_forward_unpaired.fq.gz" \
	"${sample}.trimmed2.fq.gz" "output_reverse_unpaired.fq.gz" \
	ILLUMINACLIP:TruSeq2-PE.fa:2:30:10:2:keepBothReads LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36

      rm -f output_forward_unpaired.fq.gz output_reverse_unpaired.fq.gz TruSeq2-PE.fa
      """
}

workflow.png