Long Read Processing Pipeline

A nextflow pipeline of processing long reads

Table of Contents

[Show or Hide]

1. Dependencies
2. File Format
   • Sample sheet
   • Structure of input directories
3. Usage
   • Run job
   • Usage options

Dependencies

samtools

minimap2

clair3

File Format

Sample sheet -- csv

group	barcode_start	barcode_end	barcode_template	fastq	fasta
barcode_55	1200	1237	NNNNATNNNNATNNNNATNNNNATNNNNATNNNNATNN	/path/of/fastq/directory/55	/path/of/fasta/reference_55.fa
barcode_56	1200	1237	NNNNATNNNNATNNNNATNNNNATNNNNATNNNNATNN	/path/of/fastq/directory/56	/path/of/fasta/reference_56.fa
barcode_57	1200	1237	NNNNATNNNNATNNNNATNNNNATNNNNATNNNNATNN	/path/of/fastq/directory/57	/path/of/fasta/reference_57.fa

Structure of input directories

Usage

Run job

submit the bash script below

#!/bin/bash
#BSUB -o %J.o
#BSUB -e %J.e
#BSUB -R "select[mem>1000] rusage[mem=1000]"
#BSUB -M 1000
#BSUB -q normal

# modules
module load HGI/common/nextflow/23.10.0
module load HGI/softpack/groups/team354/nf_longread
module load HGI/common/clair3

#--------------#
# user specify #
#--------------#
# LSF group
export LSB_DEFAULT_USERGROUP=hgi

# Paths
export INPUTSAMPLE=$PWD/inputs/samplesheet.csv
export OUTPUTRES=$PWD/outputs

#-----------#
# pipelines #
#-----------#
nextflow run -resume nf_longread/main.nf --sample_sheet $INPUTSAMPLE \
                                         --protocol DNA \
                                         --platform nanopore \
                                         --outdir $OUTPUTRES

Usage options

nextflow run check_inputs.nf --sample_sheet "/path/of/sample/sheet"

    Mandatory arguments:
        --sample_sheet        Path of the sample sheet
    
    Optional arguments:
    Basic:
        --outdir              the directory path of output results, default: the current directory
    
    Alignment:
        --protocol            DNA, cDNA, directRNA, default: DNA
        --platform            nanopore, pacbio, hifi, default: nanopore
    
    Variant Calling:
        --model               the trainning model of variant calling, default: ont_r10
    
    Barcode Detection:
        --mapq                the mapping quality for filtering, default: 1
        --qualcut             the base quality in the barcode for filtering , default: 10
        --numcut              the number of low-quality bases in the barcode for filtering, default: 3
        --countcut            the number of reads supporting the barcode for filtering, default: 5

    Extract SNVs:
        --basequal            the base quality for filtering, default: 30
        --region              the expected region of variants, eg: 100,200, default: 0,0

    Step arguments:
        --skip_align          skip alignment
        --skip_variant        skip variant calling
        --skip_barcode        skip barcode detection
        --skip_snvcov         skip snv coverage extraction