Illumina vs. Nanopore for Plant Pathogens — A Practical Overview

Why this comparison matters in plant pathology

High‑throughput sequencing has become a routine tool in plant pathogen research, diagnostics, and surveillance. A lot of labs are now asking: should we use Illumina or Nanopore sequencing for our plant pathogen projects? The answer isn’t one‑size‑fits‑all—it depends on your specific goals, resources, and the biology of your pathogens. For bacterial plant pathogens such as Pseudomonas, Xanthomonas, Ralstonia, Erwinia, and related taxa, the choice between Illumina short‑read sequencing and Oxford Nanopore long‑read sequencing has direct consequences for:

turnaround time
assembly quality
plasmid detection
downstream phylogenetics and diagnostics
sequencing costs

One more thing to consider is the wet-lab part, especially the extraction of gDNA. For Illumina, you can get away with a simple kit-based extraction, while for Nanopore, you really need to invest in a high-quality, high-molecular-weight DNA extraction protocol. This can be a game-changer for many labs in terms of time and resources. I tried several HMW gDNA extraction protocols (NO-MISS protocol from ONT, MagAttract HMW kit from Qiagen, Puregene, NEB Monarch HMQ gDNA kit, etc.) For now, I stick to the NEB kit which gives me consistently high yield and concentrations for gram-negative and gram-positive bacteria. So, the choice of sequencing technology also has implications for your lab workflow and budget.

When using this HMW gDNA extraction, you can achieve read N50s of around 20 kb on the MinION, which is fantastic for bacterial genomes. This means you can often get complete, closed genomes in a single run, which is a huge advantage for understanding plasmids and genomic islands. On the other hand, if you go with Illumina, you’ll get very accurate reads but your assemblies will likely be fragmented into dozens or hundreds of contigs, making it harder to resolve complex genomic regions.

For ONT library prep, I use the Rapid Barcoding Kit (SQK-RBK114.24) for its speed and simplicity, especially when processing multiple samples (i use between 12-24 samples per Minion flowcell). The Rapid kit allows you to go from DNA to library in about an hour, which is great for quick turnarounds. I also tried the Native Barcoding kit (SQK-NBD114.24) for a while, which can give you slightly longer reads and better yield, but it’s more time-consuming and requires more input DNA (400ng per sample instead of 200ng for the rapid kit). The Rapid kit is my go-to choice.

This post provides a practical, lab‑oriented comparison of Illumina and Nanopore sequencing, grounded in real Snakemake workflows used for plant‑pathogen genomics.

The technologies in one sentence

Illumina → short, highly accurate reads; ideal for SNPs, surveillance, and diagnostics
Nanopore → long, real‑time reads; ideal for complete genomes and plasmids

Short vs long reads

Read characteristics and assembly implications

Illumina (short reads)

Typical reads: 2×150 bp
Very low per‑base error rate
Assemblies are accurate but often fragmented

Best suited for:

routine diagnostics
SNP‑based phylogenetics
cgMLST / MLST
high‑throughput surveillance

Nanopore (long reads)

Read lengths: 5 kb – 100 kb+ (longest read I got was 1.2 Mb, but that’s an outlier)
Higher raw error rate (improving rapidly, especially with R10.4.1 chemistry and the new basecallers such as Dorado SUP models > version 5.0.0)
Assemblies often closed

Best suited for:

reference genome generation
plasmid and genomic island detection
structural variation
rapid outbreak investigation

Nanopore long reads

Diagnostics vs discovery

Use case	Illumina	Nanopore
Routine diagnostics	✅	⚠️
SNP resolution	✅	⚠️
Complete genomes	❌	✅
Plasmid detection	❌	✅
Turnaround time	⏳	⚡

In practice, both technologies complement each other rather than compete.

Snakemake workflows used in practice

The real impact of sequencing technology is best understood through the analysis workflows that follow.

Illumina bacterial assembly + QC workflow

Designed for high‑throughput Illumina paired‑end data, this pipeline performs:

read QC and trimming
de novo assembly
assembly QC (QUAST, BUSCO, CheckM2)
taxonomic validation
unified HTML/PDF reporting

📦 Repository
https://gitlab.ilvo.be/genomics/wgs/illumina-bacterial-assembly-snakemake

Illumina Snakemake workflow

Nanopore‑only bacterial assembly workflow

Optimized for Oxford Nanopore R10.4.1 long‑read data, focusing on:

long‑read filtering
de novo assembly with Flye
polishing and QC
reference‑grade bacterial genomes

📦 Repository
https://gitlab.ilvo.be/stevebaeyen/nanopore_only_snakemake

Nanopore Snakemake workflow

Nanopore Autocycler consensus workflow

For maximum assembly quality, this workflow integrates multiple assemblers through Autocycler and Snakemake.

Key features:

read subsampling
Flye / Canu / Plassembler consensus
QC and reporting
publication‑quality genomes

📦 Repository
https://gitlab.ilvo.be/stevebaeyen/nanopore_autocycler_snakemake

Autocycler Snakemake workflow

Assembly QC‑only workflow (technology‑agnostic)

Used when assemblies already exist (databases, Illumina, Nanopore, or hybrid).

Includes:

QUAST
BUSCO
CheckM2
ANI‑based taxonomic checks using skani/GTDB
Excel + HTML/PDF summaries

📦 Repository
https://gitlab.ilvo.be/stevebaeyen/bacterial-assembly-qc-snakemake

Assembly QC workflow

Decision tree: which pipeline should I use?

flowchart TD
    A[Start: choose a pipeline] --> B{Do you have sequencing data?}
    B -->|No| C{Main objective?}
    C -->|Diagnostics or SNPs| D[Illumina sequencing and assembly pipeline]
    C -->|Complete genome or plasmids| E[Nanopore sequencing and assembly pipeline]
    B -->|Yes| F{Data type?}
    F -->|Illumina reads| D
    F -->|Nanopore reads| G{Assembly quality needed?}
    F -->|Existing assemblies| H[Assembly QC pipeline]
    G -->|Fast and good| I[Nanopore-only pipeline]
    G -->|Reference-grade| J[Nanopore Autocycler pipeline]