workflow_w

#!/usr/bin/env nextflow

nextflow.enable.dsl=2

params.reads = "/mnt/scgc_nfs/lab/ggavelis/sandbox/sample_data/smallest_subset/*02_subseq_R{1,2}.fastq.gz"
params.outdir = "results"

Channel
    .fromFilePairs( params.reads, checkExists:true )
    .set{ read_ch }

process fastqc {
	tag "FASTQC on $sample_id"
	publishDir params.outdir

	input:
	tuple val(sample_id), path(reads)

	output:
	val sample_id, emit: sample_id
	path reads, emit: reads

	shell:
	'''
	mkdir fastqc_!{sample_id}
	fastqc -o fastqc_!{sample_id} -f fastq -q !{reads}
	'''
}

process trimmomatic {
	tag "TRIMMOMATIC on $sample_id"
	publishDir params.outdir

	input:
	val(sample_id)
	path(reads)

	output:
	val sample_id, emit: sample_id
	path "${sample_id}_trimmed_R1.fastq.gz", emit: trimmed_R1_paired
	path "${sample_id}_trimmed_R2.fastq.gz", emit: trimmed_R2_paired

	shell:
	'''
	trimmomatic \
	PE \
	-phred33 \
	-threads !{task.cpus} \
	!{reads} \
	!{sample_id}_trimmed_R1.fastq.gz \
	!{sample_id}_orphan_R1.fastq.gz \
	!{sample_id}_trimmed_R2.fastq.gz \
	!{sample_id}_orphan_R2.fastq.gz \
	LEADING:0 TRAILING:5 SLIDINGWINDOW:4:15 MINLEN:36

	'''
}

process bbmerge {
	tag "MERGING on $sample_id"
	publishDir params.outdir
	module 'singularity/3.8.0'

	cache 'lenient'
	cpus 4

	input:
	val(sample_id)
	path(trimmed_R1_paired)
	path(trimmed_R2_paired)

	output:
	path "${sample_id}_merged.fastq", emit: merged

	shell:
	'''
	reformat.sh in1=!{trimmed_R1_paired} in2=!{trimmed_R2_paired} out=!{sample_id}_merged.fastq
	'''
}


workflow {
	fastqc(read_ch)
	trimmomatic(fastqc.output.sample_id, fastqc.output.reads)
	bbmerge(fastqc.output.sample_id, trimmomatic.output.trimmed_R1_paired, trimmomatic.output.trimmed_R2_paired)
}