Prokaryotic Genome Assembly: A Step-by-Step Guide 2026

Prokaryotic genome assembly is the process of reconstructing a complete, accurate, and contiguous representation of a bacterium’s or archaeon’s DNA from fragmented sequencing data. This process is crucial for everything, from identifying novel antibiotic resistance genes to understanding microbial evolution and engineering new biotechnologies.

This roadmap provides a comprehensive, step-by-step guide to achieving high-quality prokaryotic genome assemblies, covering everything from experimental design to final validation. This roadmap is divided into key stages, each with detailed steps. I’ve included decision points and alternative approaches, as there’s no single “one-size-fits-all” solution in genomics.

1- Sample Preparation and DNA Extraction

Poor DNA integrity or contamination can lead to fragmented assemblies, sequencing biases, and misleading biological conclusions. Follow these steps to maximize DNA quality:

DNA Extraction

Use HMW DNA kits: Prioritize kits designed for high-molecular-weight (HMW) DNA extraction (e.g., Qiagen Genomic-tip, MagAttract HMW DNA Kit). Also, you can use Alternative methods such as Phenol-chloroform extraction, but ensure proper phase separation and avoid RNA/protein carryover.

Quality Control (QC):

1. 1. Quantify and assess purity:

• 1. • Nanodrop: Check for contaminants (ideal ratios: A260/A280 ≈ 1.8–2.0; A260/A230 > 2.0).
     • Qubit: For accurate concentration (fluorometry avoids overestimation by Nanodrop).
  2. Visualize DNA integrity:
     • Agarose gel electrophoresis: Look for a tight, high-molecular-weight band (>20 kb) with minimal smearing (indicates shearing).
     • Pulse-field gel electrophoresis (PFGE): Gold standard for confirming HMW DNA (>50 kb).

Troubleshooting Table

Tip: If your DNA fails QC, do not proceed to sequencing—repeat extraction.

2- Sequencing Technology Selection

Prokaryotic genome assembly hinges on sequencing technology choice. While short-read data (Illumina) excels at accuracy, long-read data (PacBio/Nanopore) resolves repetitive regions and structural complexity. Combining both (hybrid assembly) is the gold standard for closed, high-quality genomes. Below, we break down options for every budget and goal.

1- Short-Read Sequencing (e.g., Illumina)

• Platforms: NovaSeq (high throughput), MiSeq (rapid turnaround), NextSeq (mid-scale projects).
• High accuracy (~99.9%)
• Low error rate
• Ideal for detecting small variants (SNPs, indels)
• Limitations: Struggles with repetitive regions and large structural variations

2- Long-Read Sequencing (e.g., PacBio, Oxford Nanopore Technologies (ONT)):

• Generates ultra-long reads (>50 kb)
• Essential for resolving repeats, plasmids, and structural variations
• PacBio HiFi: High accuracy (~99.9%)
• Nanopore: Longer reads but higher error rate (~95%)

Best Practice: Hybrid sequencing (Illumina + PacBio HiFi) provides the best balance of accuracy and completeness.

Sequencing Platform Comparison

3- Data Preprocessing & Quality Control:

 Raw Data Inspection

• • Tools: FastQC (for short reads), NanoPlot/PycoQC (for long reads).
  • Metrics to Examine:
    • Read length distribution.
    • Base quality scores (Phred scores).
    • GC content.
    • Presence of adapter sequences.
    • Overrepresented sequences (potential contamination).
  • Visualization: Generate plots to visualize these metrics. This helps identify potential problems early on.

Trimming and Filtering

Remove low-quality bases, adapter sequences, and short reads. This improves assembly accuracy and reduces computational burden. The best Tools for Trimming and Filtering:

• • • • Short Reads: Trimmomatic, Cutadapt, Sickle.
      • Long Reads: Canu (includes trimming), Filtlong, Porechop (for ONT).

Error Correction for Long Reads (Optional)

Reduce the error rate of long reads, particularly for ONT data. There are different methods, such as:

• • • Self-Correction: Use tools like Canu, MECAT, or Falcon, which align long reads to each other to correct errors.
    • Hybrid Correction: Use short, accurate reads to correct long reads. Tools: Racon, Pilon, Medaka (ONT-specific).

Note: PacBio HiFi reads generally do not require error correction.

4- Genome Assembly Approaches

After rigorous QC, the next challenge is reconstructing the genome from sequencing reads. The choice of assembler and strategy depends on your data type (short-read, long-read, or hybrid) and biological goals (draft vs. closed genome). Below, we break down the three main assembly paradigms:

1- De Bruijn Graph Assemblers (Short Reads Only):

Principle: Breaks reads into k-mers (small overlapping fragments) to build paths through a graph.

Best For: Illumina-only data

Tools:

• SPAdes: Optimized for bacterial genomes. Use --isolate mode for single strains.
• Velvet/ABySS: Older tools, but useful for benchmarking.

Pros:

• • Fast for small datasets.
  • Effective for low-complexity regions.

Cons:

• Fragmented assemblies due to repeats.
• Struggles with plasmids.

2- Overlap-Layout-Consensus (OLC) Assemblers (Long Reads):

Principle: Uses overlaps between long reads to build contigs.

Best For: PacBio/Nanopore data. Ideal for resolving repeats and circularizing genomes.

Tools:

• • Flye: Fast, supports Nanopore/PacBio. Use --meta for metagenomes.
  • Canu: Robust but computationally heavy. Great for noisy long reads.
  • Shasta: Ultra-fast for Nanopore (optimized for human genomes but adaptable).

Pros:

• • Produces near-complete chromosomes/plasmids.
  • Handles repetitive regions (e.g., rRNA operons).

Cons:

• • Requires high-quality long reads.
  • Computationally intensive.

3- Hybrid Assemblers (Short + Long Reads):

Principle: Combines short-read accuracy with long-read contiguity.

Best For: Closed, polished genomes. Gold standard for prokaryotes.

Tools:

• • Unicycler: Specifically designed for bacteria. Automates hybrid assembly and polishing.
  • MaSuRCA: Hybrid + meta-assembly support.
  • OPERA-MS: For complex/mixed communities.

5- Assembly Evaluation

Once you’ve run your genome assembler, you’re not quite done! A crucial step is to assess the quality of your assembly. Just like building with LEGOs, you want to know if you’ve created a complete, accurate, and well-structured representation of the original genome. Several key metrics help us evaluate this, falling broadly into categories of contiguity, completeness & contamination, and accuracy & biological consistency. Let’s explore these:

1- Contiguity

Contiguity measures how connected or unfragmented your assembly is. Think of it like having fewer puzzle pieces and larger chunks put together. Higher contiguity generally indicates a better assembly, making downstream analyses like gene annotation and comparative genomics more reliable.

• N50: This refers to the length of the shortest contig among the longest contigs that collectively cover 50% of the genome. Aim for an N50 greater than 100 kb.
• L50: Number of contigs required to cover 50% of the genome. Ideal scenario: L50 = 1 (chromosome) + plasmids.

Tip: For a good prokaryotic genome assembly, you should ideally aim for a low number of contigs, ideally in the range of 1–5 contigs. This often reflects a single chromosome and potentially a few plasmids, which are common in bacteria.

2. Completeness & Contamination

A high-quality genome assembly should be nearly complete while minimizing contamination from other organisms or misassembled sequences.

1. CheckM Analysis
   • Completeness: Measures the presence of conserved bacterial marker genes.
   • Target: >95% completeness for a high-quality assembly.
   • Contamination: Detects extra copies of marker genes, indicating possible contamination or disassembly.
   • Acceptable threshold: <5% contamination.

3. Accuracy & Biological Consistency

Genome assemblies should match biological expectations in terms of size and composition. The total genome size should be within ±10% of the expected size based on related species. Additionally, the genome’s GC content should match the known GC% of the species (within ±1–2%).

6. Genome Annotation:

After assembling a high-quality prokaryotic genome, the next critical step is gene prediction and functional annotation. This process identifies the locations of protein-coding genes, tRNA genes, and rRNA genes, and assigns functional roles to predicted proteins based on sequence similarity and conserved domains. The Recommended Tools for Genome Annotation:

• Prokka (Highly Recommended): A fast and comprehensive prokaryotic genome annotation pipeline that integrates multiple gene prediction tools. It efficiently identifies protein-coding genes, tRNA genes, and rRNA genes.

• NCBI Prokaryotic Genome Annotation Pipeline (PGAP): PGAP is the annotation pipeline used by the National Center for Biotechnology Information (NCBI) for annotating all prokaryotic genomes submitted to their databases like GenBank. Using PGAP ensures your annotation is consistent with the standards used by a major public repository. You can submit your genome to NCBI for annotation through their online submission portal, or in some cases, download and run PGAP locally (though local setup can be more complex).

Conclusion

Prokaryotic genome assembly is a powerful gateway to understanding the microbial world. From unraveling the complexities of microbiomes to combating antibiotic resistance, the insights gained from assembled genomes are invaluable. While challenges exist – from repetitive regions to data quality – a fantastic suite of bioinformatics tools is available to help you overcome these hurdles.

In this post, we’ve explored some of the best bioinformatics apps for prokaryotic genome assembly, including SPAdes, Unicycler, Flye, and Canu. The “best” choice truly depends on your data type (short reads, long reads, or both), your computational resources, and your research goals.

Have questions or tips? Share your genome assembly experiences below!

References:

1. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017 Jun 8;13(6):e1005595.
2. Torsten Seemann, Prokka: rapid prokaryotic genome annotation, Bioinformatics, Volume 30, Issue 14, July 2014, Pages 2068–2069, https://doi.org/10.1093/bioinformatics/btu153
3. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 2014 Aug 1;30(15):2114-20. doi: 10.1093/bioinformatics/btu170.