ONTViSc (ONT-based Viral Screening for Biosecurity)

Introduction

eresearchqut/ontvisc is a Nextflow-based bioinformatics pipeline designed to help diagnostics of viruses and viroid pathogens for biosecurity. It takes fastq files generated from either amplicon or whole-genome sequencing using Oxford Nanopore Technologies as input.

The pipeline can either: 1) perform a direct search on the sequenced reads, 2) generate clusters, 3) assemble the reads to generate longer contigs or 4) directly map reads to a known reference.

The reads can optionally be filtered from a plant host before performing downstream analysis.

Pipeline overview

• Data quality check (QC) and preprocessing
  • Merge fastq files (Fascat, optional)
  • Raw fastq file QC (Nanoplot)
  • Trim adaptors (PoreChop ABI - optional)
  • Filter reads based on length and/or quality (Chopper - optional)
  • Reformat fastq files so read names are trimmed after the first whitespace (bbmap)
  • Processed fastq file QC (if PoreChop and/or Chopper is run) (Nanoplot)
• Host read filtering
  • Align reads to host reference provided (Minimap2)
  • Extract reads that do not align for downstream analysis (seqtk)
• QC report
  • Derive read counts recovered pre and post data processing and post host filtering
• Read classification analysis mode
• Clustering mode
  • Read clustering (Rattle)
  • Convert fastq to fasta format (seqtk)
  • Cluster scaffolding (Cap3)
  • Megablast homology search against ncbi or custom database (blast)
  • Derive top candidate viral hits
  • Align reads back to top reference and derive coverage statistics (mosdepth and coverM)
• De novo assembly mode
  • De novo assembly (Canu or Flye)
  • Megablast homology search against ncbi or custom database or reference (blast)
  • Derive top candidate viral hits
  • Align reads back to top reference and derive coverage statistics (mosdepth and coverM)
• Read classification mode
  • Option 1 Nucleotide-based taxonomic classification of reads (Kraken2, Braken)
  • Option 2 Protein-based taxonomic classification of reads (Kaiju, Krona)
  • Option 3 Convert fastq to fasta format (seqtk) and perform direct homology search using megablast (blast)
• Map to reference mode
  • Align reads to reference fasta file (Minimap2) and derive bam file and alignment statistics (Samtools)

Code and detailed instructions can be found here. A comprehensive, step-by-step guide on setting up and executing the ONTViSc pipeline across three high-performance computing systems hosted by Australian research and computing facilities - Lyra (Queensland University of Technology), Gadi (National Computational Infrastructure), and Setonix (Pawsey) - utilising the Australian Nextflow Seqera Service, can be found here.

Authors

Marie-Emilie Gauthier
Craig Windell
Magdalena Antczak
Roberto Barrero