How to Choose a Long-Read Sequencing Platform: PacBio vs Oxford Nanopore

Long-read sequencing platforms from Pacific Biosciences (PacBio) and Oxford Nanopore Technologies (ONT) have changed how researchers approach genome assembly, variant detection, and epigenetic analysis. This article compares the two main long-read sequencing technologies across throughput, read length, accuracy, cost, scalability, and real-time analysis to help you select a platform based on your specific project needs. The decision between PacBio and ONT depends on your priorities: PacBio generally offers higher single-molecule accuracy with circular consensus sequencing, while ONT provides higher throughput, lower upfront instrument costs, and real-time data streaming. For most research applications, the choice comes down to whether you need maximum accuracy per base or maximum data volume per dollar.

Understanding Long-Read Sequencing Technologies

Long-read sequencing technologies produce DNA sequence reads that are substantially longer than the 150 to 300 base pair reads generated by short-read platforms like Illumina. These longer reads allow researchers to span repetitive regions, resolve structural variants, and generate more contiguous genome assemblies. Third-generation sequencing techniques have become increasingly popular because they can produce long, high-quality reads that overcome many limitations of short-read approaches [6].

The two dominant long-read platforms operate on fundamentally different principles. PacBio uses single-molecule real-time (SMRT) sequencing, where DNA polymerase incorporates fluorescently labeled nucleotides into a growing strand while the signal is detected in real time. ONT uses nanopore sequencing, where a DNA molecule passes through a protein pore embedded in a membrane, and the changes in electrical current as nucleotides pass through the pore are converted into sequence information.

Both platforms have evolved significantly since their introduction. PacBio has developed circular consensus sequencing (CCS) that reads the same molecule multiple times to generate highly accurate consensus sequences. ONT has improved its pore chemistry and basecalling algorithms to increase accuracy while maintaining its distinctive real-time data streaming capability.

At a Glance: Platform Comparison

The following table summarizes the key differences between PacBio and ONT platforms for common research applications.

A comparative evaluation of PacBio and ONT for 16S rRNA-based soil microbiome profiling found that both platforms provided comparable bacterial diversity assessments, with PacBio showing slightly higher efficiency in detecting low-abundance taxa [10]. Despite differences in sequencing accuracy, ONT produced results that closely matched those of PacBio, suggesting that ONT inherent sequencing errors do not significantly affect the interpretation of well-represented taxa [10].

Core Principles of Platform Selection

Read Length and Fragment Size Bias

Both PacBio and ONT platforms show biases toward sequencing longer DNA fragments when presented with a mixture of different fragment sizes. A study evaluating both platforms for analysis of long cell-free DNA in plasma found that both showed biases to sequence longer fragments (1500 base pairs versus 200 base pairs), with PacBio showing a stronger bias with a 5-fold overrepresentation of long fragments compared to a 2-fold bias in ONT [5]. This size bias has practical implications for projects analyzing cell-free DNA, where fragment size distribution carries biological information.

For projects focused on short tandem repeats (STRs), long-read sequencing enables direct sequencing of full STR regions with improved accuracy compared to short-read approaches [9]. Traditional short-read sequencing struggles to accurately characterize STRs due to limited read length, which limits the ability to resolve repeat expansions, increases mapping errors, and reduces sensitivity for detecting large insertions or interruptions [9]. Both ONT and PacBio can overcome these limitations, but the choice depends on whether you need the higher accuracy of PacBio circular consensus sequencing or the higher throughput of ONT.

Accuracy Considerations

Accuracy differs between the platforms in ways that matter for different applications. PacBio circular consensus sequencing achieves high accuracy by reading the same molecule multiple times. A study combining unique molecular identifiers with ONT or PacBio circular consensus sequencing found that to reach a mean UMI consensus error rate below 0.01 percent, a UMI read coverage of 15 times was needed for ONT R10.3, 25 times for ONT R9.4.1, and 3 times for PacBio circular consensus sequencing [12]. The resulting mean error rates were 0.0042 percent for ONT R10.3, 0.0041 percent for ONT R9.4.1, and 0.0007 percent for PacBio [12].

For single nucleotide variant detection, ONT long-read sequencing demonstrated competitive yet slightly lower accuracy than short-read sequencing in high-complexity regions, with an F-measure of 0.954 compared to 0.968 for short-read sequencing [11]. For indel detection, ONT showed robust performance for small indels of 1 to 5 base pairs in high-complexity regions with an F-measure of 0.869, but accuracy decreased significantly in low-complexity regions and for larger indels [11]. ONT identified 2.86 times more structural variants than short-read sequencing, with superior detection of large-scale variations [11].

Throughput and Data Volume

Throughput differences between the platforms can determine which one is practical for your project. The study of long cell-free DNA found that while PacBio generated data with higher percentages of long cell-free DNA fragments, a higher number of long cell-free DNA fragments eligible for tissue-of-origin analysis could be obtained from nanopore sequencing due to its much higher throughput [5]. This throughput advantage matters when your downstream analysis requires a minimum number of fragments or reads to achieve statistical power.

For genome assembly projects, the choice of platform affects assembly quality and completeness. A comparison of genome assembly tools for Mycobacterium tuberculosis using hybrid sequencing technologies found that Unicycler-based assemblies had significantly higher genome completeness at approximately 98.7 percent compared to other assembler tools [7]. The number of genes from hybrid assemblies with ONT and PacBio long reads was greater than short-read assembly alone, with a mean of 4,620 genes compared to 4,478 genes for short-read assembly alone [7].

Practical Workflow Considerations

Sample Preparation and Input Requirements

Sample preparation differs between the platforms in terms of DNA quantity, quality, and fragmentation requirements. PacBio typically requires higher molecular weight DNA with specific quality metrics, while ONT has developed protocols that work with a range of input quantities and quality levels. Both platforms require careful attention to DNA purity, as contaminants can affect enzyme activity in PacBio sequencing or pore performance in ONT sequencing.

For projects working with challenging sample types such as cell-free DNA, the pre-analytical variability is a critical consideration. Long-read sequencing for cancer liquid biopsy faces technical and translational barriers including pre-analytical variability, cost, and high computational demands [8]. When planning a project with limited or degraded samples, you should evaluate whether the platform can work with your input material or whether additional optimization is needed.

Sequencing Run Design

Run design decisions include the number of samples to multiplex, the sequencing depth required, and the expected read length distribution. Sequencing depth strongly influences variant calling performance across all variant types for ONT long-read sequencing, while multiplexing effects are minimal after controlling for depth [11]. This finding suggests that you should prioritize sequencing depth over multiplexing considerations when designing your experiment.

For amplicon sequencing projects, the choice of platform affects the accuracy of consensus sequences. The UMI-based approach combining unique molecular identifiers with ONT or PacBio sequencing achieved chimera rates below 0.02 percent for ribosomal RNA operon amplicons of approximately 4,500 base pairs and genomic sequences greater than 10,000 base pairs [12]. If your project requires high-accuracy consensus sequences for targeted regions, the UMI approach can be used with either platform, but the required coverage differs substantially.

Real-Time Analysis and Data Streaming

ONT platforms offer real-time data streaming, which allows you to monitor sequencing progress and make decisions during the run. This capability is valuable for applications where you need to stop sequencing once sufficient data has been collected or where rapid results are needed. PacBio platforms typically require the full sequencing run to complete before data analysis can begin.

Real-time analysis also enables adaptive sampling, where the sequencing device can selectively reject or accept reads based on their characteristics. This feature is unique to ONT and can be used to enrich for specific genomic regions without additional sample preparation.

Options and Tradeoffs

PacBio Platform Options

PacBio offers several instrument models with different throughput and cost profiles. The choice of instrument depends on your project scale and budget. Smaller instruments are suitable for targeted sequencing and amplicon projects, while larger instruments provide the throughput needed for whole-genome sequencing at scale.

PacBio circular consensus sequencing is particularly well suited for applications requiring high accuracy, such as variant confirmation, amplicon sequencing, and methylation analysis. The higher accuracy per base comes at the cost of lower throughput compared to ONT, as each molecule must be read multiple times to generate the consensus sequence.

ONT Platform Options

ONT offers a range of devices from portable options suitable for field sequencing to high-throughput instruments for large-scale projects. The PromethION platform provides the highest throughput and is appropriate for whole-genome sequencing projects. The portable devices enable sequencing in remote locations or clinical settings where traditional laboratory infrastructure is not available.

ONT flow cells can be purchased individually, which provides flexibility for projects with variable sequencing needs. The ability to run one flow cell at a time or multiple flow cells in parallel allows you to scale sequencing capacity to match your project requirements.

Hybrid Approaches

Hybrid sequencing approaches combine long-read data with short-read data to leverage the strengths of both technologies. The comparison of genome assembly tools for Mycobacterium tuberculosis found that hybrid assemblies with ONT and PacBio long reads detected more genes than short-read assembly alone [7]. Pan-genome analysis of Illumina and PacBio hybrid assemblies revealed the greatest number of detected genes at 4,639 genes compared to the H37Rv reference containing 3,976 genes [7].

Hybrid approaches are particularly valuable for genome assembly projects where the long reads provide contiguity and the short reads provide accuracy. The choice of which long-read platform to use in a hybrid approach depends on the specific requirements of your project, including the size of the genome, the complexity of the repetitive content, and your budget.

Observations and Measurements

Fragment Size Distribution Analysis

When analyzing cell-free DNA, the fragment size distribution carries biological information that can be affected by the sequencing platform. The study comparing PacBio and ONT for long cell-free DNA analysis found that percentages of cell-free DNA fragments of 500 base pairs were around 6-fold higher in PacBio compared to ONT [5]. This difference in observed fragment size distribution has implications for studies that use fragment size as a biomarker or for tissue-of-origin analysis.

End motif profiles of cell-free DNA from PacBio and ONT were similar, yet exhibited platform-dependent patterns [5]. If your project involves end motif analysis, you should be aware that the platform you choose may introduce systematic biases that need to be accounted for in your analysis.

Methylation and Epigenetic Analysis

Both platforms can detect DNA methylation through different mechanisms. PacBio detects methylation through polymerase kinetics, while ONT detects modified bases through changes in electrical current. Tissue-of-origin analysis based on single-molecule methylation patterns showed comparable performance on both platforms in the cell-free DNA study [5].

For projects requiring methylation analysis, the choice of platform may depend on the specific methylation marks you need to detect and the throughput required. The Giraffe tool was developed to facilitate assessment of genomic regional methylation proportions for both DNA and direct RNA sequencing reads across different platforms [6].

Structural Variant Detection

Long-read sequencing excels at detecting structural variants that are difficult or impossible to resolve with short-read sequencing. ONT identified 2.86 times more structural variants than short-read sequencing, with superior detection of large-scale variations [11]. This advantage is particularly relevant for cancer genomics, where structural variants play a significant role in tumor biology.

Long-read sequencing technologies, including both SMRT and nanopore sequencing, offer opportunities to overcome limitations of short-read sequencing by preserving long-range molecular information and enabling multimodal characterization of tumor-derived material in biofluids [8]. For cancer liquid biopsy applications, long-read sequencing has shown promise in detecting structural variants, methylation patterns, and tumor-of-origin signals that may not be fully captured by short-read approaches [8].

Records and Documentation

Sequencing Run Records

Maintain detailed records of each sequencing run, including the platform, instrument model, flow cell or chip lot number, reagent lot numbers, sample preparation protocol, and run parameters. These records are essential for troubleshooting, reproducibility, and quality control. When comparing data across platforms or across runs, documentation of run conditions allows you to identify sources of variation.

Record the following metrics for each run: read length distribution, read quality scores, throughput in bases or reads, and any platform-specific quality metrics. For PacBio, record polymerase read length and number of passes for circular consensus sequencing. For ONT, record pore occupancy, read quality scores, and basecalling model version.

Quality Control Metrics

Quality control metrics should be tracked consistently across runs and platforms. The Giraffe tool was developed to facilitate comparative analysis and visualization across diverse samples and platforms, enabling assessment of read quality, sequencing bias, and genomic regional methylation proportions [6]. Using standardized tools for quality assessment allows you to compare data quality across platforms and identify potential issues early in the analysis pipeline.

For projects comparing multiple samples or platforms, effective comparative analysis is essential for understanding biological mechanisms and establishing benchmark baselines [6]. Without comprehensive tools for data comparison and visualization, researchers with limited bioinformatics experience face challenges in interpreting their data [6].

Common Failure Patterns

Size Bias Misinterpretation

A common failure pattern is interpreting observed fragment size distributions without accounting for platform-specific size biases. Both PacBio and ONT show biases toward sequencing longer fragments, but the magnitude of the bias differs between platforms [5]. If you are comparing fragment size distributions across platforms or interpreting fragment size as a biological signal, you must account for these platform-specific biases.

Insufficient Sequencing Depth

Another common failure is underestimating the sequencing depth required for your application. Sequencing depth strongly influences variant calling performance across all variant types for ONT long-read sequencing [11]. For projects requiring high-accuracy consensus sequences, the required coverage depends on the platform and pore version, with ONT requiring substantially higher coverage than PacBio for the same accuracy level [12].

Inadequate Bioinformatics Resources

Long-read sequencing generates large amounts of data that require substantial computational resources for analysis. High computational demands are a recognized barrier to the adoption of long-read sequencing in clinical applications [8]. Before selecting a platform, assess whether you have the computational infrastructure and bioinformatics expertise to process and analyze the data you will generate.

Ignoring Platform-Specific Error Profiles

Each platform has distinct error profiles that affect downstream analysis. ONT sequencing errors do not significantly affect the interpretation of well-represented taxa in microbiome profiling, but they may affect detection of low-abundance taxa [10]. Understanding the error profile of your chosen platform is essential for designing appropriate analysis pipelines and interpreting results.

Limitations and Considerations

Accuracy Limitations

While both platforms have improved accuracy over time, they still have limitations compared to short-read sequencing for certain variant types. ONT long-read sequencing showed competitive yet slightly lower accuracy than short-read sequencing for single nucleotide variant detection in high-complexity regions [11]. For indel detection, accuracy decreased significantly in low-complexity regions and for larger indels [11].

PacBio circular consensus sequencing achieves higher accuracy but requires multiple passes of the same molecule, which reduces throughput. The tradeoff between accuracy and throughput is a fundamental consideration in platform selection.

Cost Considerations

Cost structures differ substantially between the platforms. ONT has lower upfront instrument costs, making it more accessible for individual laboratories or smaller projects. PacBio instruments have higher upfront costs but may offer lower per-base costs for high-accuracy applications. The total cost of a project includes instrument purchase or access, consumables, personnel time, and computational resources.

For large-scale projects, the higher throughput of ONT may provide a cost advantage despite lower per-read accuracy. For projects requiring maximum accuracy, the higher cost of PacBio may be justified by the reduced need for additional sequencing or validation.

Computational Demands

Both platforms generate data that require substantial computational resources for basecalling, alignment, and variant calling. ONT basecalling can be performed in real time during sequencing, but this requires appropriate computational hardware. PacBio data analysis typically requires high-performance computing resources for genome assembly and variant detection.

High computational demands are a recognized barrier to the broader adoption of long-read sequencing [8]. When planning a project, include computational costs in your budget and ensure that your institution has the necessary infrastructure.

Safety and Regulatory Context

Data Sharing and Privacy

Genomic data generated by long-read sequencing platforms may be subject to data sharing policies and privacy regulations. The National Institutes of Health Genomic Data Sharing Policy provides expectations for the sharing of genomic data generated through NIH-funded research [3]. If your research is funded by NIH or involves human subjects, you should review the applicable data sharing requirements before selecting a platform and generating data.

For projects involving human samples, consider the privacy implications of generating long-read sequencing data. Long reads can span multiple genetic variants, potentially increasing the identifiability of samples. Ensure that your data management plan addresses these considerations.

Data Management and Reproducibility

The FAIR Guiding Principles provide a framework for making data findable, accessible, interoperable, and reusable [4]. When selecting a sequencing platform, consider how the data formats and analysis outputs align with FAIR principles. Documentation of analysis pipelines and parameters is essential for reproducibility.

The EMBL-EBI Training program offers resources for researchers seeking to develop bioinformatics skills for analyzing sequencing data [1]. The NCBI Data Resources provide access to public databases for depositing and accessing genomic data [2]. Familiarize yourself with these resources to ensure that your data management practices align with community standards.

Professional Escalation Criteria

When to Consult a Bioinformatics Specialist

If your project involves complex genome assembly, structural variant detection, or integration of multiple data types, consult a bioinformatics specialist before selecting a platform. The choice of assembly tools and analysis pipelines can significantly affect results, as demonstrated by the comparison of genome assembly tools for Mycobacterium tuberculosis [7]. A specialist can help you design an appropriate analysis strategy and avoid common pitfalls.

When to Seek Technical Support

If you encounter unexpected results in sequencing quality metrics, read length distributions, or throughput, contact the platform manufacturer technical support. Both PacBio and ONT provide technical support for their instruments and protocols. Early intervention can prevent wasted time and resources.

When to Reconsider Platform Choice

If your project requirements change during the course of your research, reconsider whether your chosen platform remains appropriate. For example, if you initially planned a targeted sequencing project but now need whole-genome sequencing, the throughput and cost characteristics of your platform may no longer be optimal. Similarly, if you discover that your application requires higher accuracy than initially anticipated, you may need to switch to a platform with higher accuracy or implement additional error correction strategies.

Frequently Asked Questions

What is the main difference between PacBio and Oxford Nanopore sequencing?

PacBio uses single-molecule real-time sequencing with DNA polymerase and fluorescent nucleotides, while Oxford Nanopore uses nanopore electrical current detection as DNA passes through a protein pore. PacBio circular consensus sequencing achieves higher accuracy by reading the same molecule multiple times, while ONT offers higher throughput and real-time data streaming. Both platforms produce long reads that enable analysis of repetitive regions and structural variants that are difficult to resolve with short-read sequencing [6][9].

Which platform has higher accuracy for single nucleotide variant detection?

PacBio circular consensus sequencing generally provides higher per-base accuracy than ONT single-pass sequencing. A study combining unique molecular identifiers with both platforms found that PacBio required only 3 times coverage to reach a mean consensus error rate below 0.01 percent, while ONT required 15 to 25 times coverage depending on the pore version [12]. However, ONT accuracy has improved with newer pore versions and basecalling algorithms, and the practical impact of accuracy differences depends on your specific application.

How does sequencing depth affect variant calling performance?

Sequencing depth strongly influences variant calling performance across all variant types for ONT long-read sequencing [11]. Higher depth improves sensitivity and precision for detecting single nucleotide variants, indels, and structural variants. When designing your experiment, determine the minimum depth required for your application and account for the platform-specific error rates in your depth calculation.

Can I use both PacBio and ONT in the same project?

Yes, hybrid approaches that combine long-read data with short-read data are increasingly popular. A comparison of genome assembly tools for Mycobacterium tuberculosis found that hybrid assemblies with ONT and PacBio long reads detected more genes than short-read assembly alone [7]. You can also use both long-read platforms in the same project, though this increases cost and complexity. The choice depends on whether the complementary strengths of the platforms justify the additional expense.

Which platform is better for analyzing cell-free DNA?

The choice depends on your specific analysis goals. PacBio generates data with higher percentages of long cell-free DNA fragments, but ONT provides a higher number of long cell-free DNA fragments eligible for tissue-of-origin analysis due to its much higher throughput [5]. Both platforms show biases toward sequencing longer fragments, with PacBio showing a stronger bias [5]. Consider whether you need maximum fragment length information or maximum number of analyzable fragments.

How do the platforms compare for microbiome profiling?

A comparative evaluation of 16S rRNA gene sequencing found that ONT and PacBio provided comparable bacterial diversity assessments, with PacBio showing slightly higher efficiency in detecting low-abundance taxa [10]. Despite differences in sequencing accuracy, ONT produced results that closely matched those of PacBio, suggesting that ONT inherent sequencing errors do not significantly affect the interpretation of well-represented taxa [10]. Both platforms enabled clear clustering of samples based on soil type [10].

What are the main cost differences between the platforms?

ONT generally has lower upfront instrument costs, making it more accessible for individual laboratories or smaller projects. PacBio instruments have higher upfront costs but may offer advantages for high-accuracy applications. The total cost of a project includes instrument purchase or access, consumables, personnel time, and computational resources. For large-scale projects, the higher throughput of ONT may provide a cost advantage despite lower per-read accuracy.

How should I choose between PacBio and ONT for my project?

Start by defining your project requirements: the type of variants you need to detect, the accuracy required, the throughput needed, your budget, and your computational resources. If you need maximum accuracy for targeted regions, PacBio circular consensus sequencing may be appropriate. If you need high throughput for large-scale projects or real-time analysis, ONT may be better suited. Consider running a small pilot study with both platforms to evaluate performance on your specific sample types before committing to a large-scale project.

Related Bioinformatics Guides

• Long-Read Sequencing Technologies: PacBio and Oxford Nanopore
• Long Read Metagenomic Assembly: Structural Analysis and Computational Methodologies in Bioinformatics
• Basecalling Algorithms for Nanopore Sequencing
• Long-Read Genome Assembly and Polishing Strategies
• Nanopore Adaptive Sampling for Targeted Pathogen Sequencing

References and Further Reading

• EMBL-EBI Training. European Bioinformatics Institute.
• NCBI Data Resources. National Center for Biotechnology Information.
• Genomic Data Sharing Policy. National Institutes of Health.
• The FAIR Guiding Principles. Scientific Data.
• Comparison of Single Molecule, Real-Time Sequencing and Nanopore Sequencing for Analysis of the Size, End-Motif, and Tissue-of-Origin of Long Cell-Free DNA in Plasma.. Clinical chemistry, 2023.
• Giraffe: A tool for comprehensive processing and visualization of multiple long-read sequencing data.. Computational and structural biotechnology journal, 2024.
• Comparisons of genome assembly tools for characterization of Mycobacterium tuberculosis genomes using hybrid sequencing technologies.. PeerJ, 2024.
• Long-read sequencing for cancer liquid biopsy: advancing precision oncology.. 2026.
• Get ready for short tandem repeats analysis using long reads-the challenges and the state of the art. Frontiers in Genetics, 2025.
• Comparative evaluation of sequencing platforms: Pacific Biosciences, Oxford Nanopore Technologies, and Illumina for 16S rRNA-based soil microbiome profiling. Frontiers in Microbiology, 2025.
• Investigating the Performance of Oxford Nanopore Long-Read Sequencing with Respect to Illumina Microarrays and Short-Read Sequencing. bioRxiv, 2025.
• High-accuracy long-read amplicon sequences using unique molecular identifiers with Nanopore or PacBio sequencing. Nature Methods, 2021.

This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.