Troubleshooting Common Problems in Long-Read Sequencing of Bacterial Genomes

By Dr. Zubair Khalid, DVM, MS, PhD · 2026-08-21

Quick Answer

• Diagnose long-read sequencing failures by checking DNA quality, library preparation, sequencing conditions, and basecalling parameters in that order.
• Measure DNA purity, concentration, and fragment length before loading a flowcell, since contaminated or degraded input DNA is the most common cause of low yield.
• Recognize that even with optimal preparation, some samples resist improvement, and pore loss or underperformance may require repeating the run instead of adjusting software settings.

At a Glance

Understanding Long-Read Sequencing for Bacterial Genomes

Long-read sequencing platforms such as Oxford Nanopore and PacBio generate reads that span thousands to hundreds of thousands of base pairs. This read length is the primary advantage over short-read technologies because it allows researchers to resolve repetitive regions, complete circular chromosomes, and assemble genomes with fewer contigs. For bacterial genomics, this capability supports tasks such as identifying biosynthetic gene clusters, tracking antimicrobial resistance determinants, and performing pathogen surveillance.

The workflow for long-read sequencing of bacterial genomes follows a standard sequence of steps. First, you extract high-molecular-weight DNA from a pure culture or clinical sample. Second, you prepare a sequencing library by attaching adapters and, in some cases, amplifying the DNA. Third, you load the library onto a flowcell and initiate the sequencing run. Fourth, you basecall the raw signal data to produce nucleotide sequences. Finally, you assemble the reads into a complete genome and perform downstream analysis.

Failures can occur at any of these stages. A systematic troubleshooting approach requires you to isolate the variable that changed between a successful run and a failed run. Keep detailed records of each run, including DNA quality metrics, library preparation conditions, flowcell lot numbers, and basecalling parameters. These records allow you to compare runs and identify the factor that correlates with the failure.

The two dominant long-read platforms differ in their chemistry and failure modes. Oxford Nanopore sequencing passes DNA through a protein pore and measures changes in electrical current as nucleotides pass through. PacBio sequencing uses a zero-mode waveguide to observe DNA synthesis in real time. Each platform has distinct requirements for DNA quality, library preparation, and basecalling. Your troubleshooting strategy should account for these differences.

DNA Quality Assessment and Extraction

Measuring DNA Purity and Concentration

The quality of input DNA is the single most important factor in long-read sequencing success. Contaminants such as proteins, polysaccharides, salts, and ethanol can interfere with library preparation enzymes and block nanopores. Measure DNA purity using spectrophotometry, which reports absorbance ratios at 260 and 280 nanometers for protein contamination and at 260 and 230 nanometers for salt or carbohydrate contamination. A ratio near 1.8 at 260/280 and above 2.0 at 260/230 generally indicates acceptable purity, though these values vary by organism and extraction method.

Fluorometry provides a more accurate measurement of double-stranded DNA concentration than spectrophotometry because it specifically binds to double-stranded DNA instead of measuring all nucleic acids. Use a fluorometer with a DNA-binding dye to quantify your sample before proceeding to library preparation. This measurement is critical because overloading or underloading the library preparation reaction can reduce yield and read quality.

For bacterial samples, the extraction method should be matched to the cell wall structure of the organism. Gram-positive bacteria require more vigorous lysis than Gram-negative bacteria. Enzymatic digestion with lysozyme or mutanolysin may be necessary for organisms with thick peptidoglycan layers. Mycobacteria and other organisms with waxy cell walls require additional mechanical disruption or chemical treatment. The American Type Culture Collection provides authenticated strain information and guidance on culture conditions that can help you standardize your starting material.

Assessing DNA Fragment Length

Long-read sequencing requires high-molecular-weight DNA. Fragmented DNA produces short reads that negate the advantage of the platform. Assess fragment length using pulsed-field gel electrophoresis, a TapeStation, or a Femto Pulse system. The acceptable fragment size distribution depends on the sequencing platform and the read length you need for your assembly goal.

For complete bacterial genome assembly, you generally want DNA fragments that exceed the length of the repetitive elements you need to resolve. The 16S ribosomal RNA operon, insertion sequences, and transposons can be several kilobases long. If your DNA fragments are shorter than these repeats, the assembler cannot bridge them, and your genome will contain gaps or misassemblies.

Handle high-molecular-weight DNA gently during extraction and library preparation. Avoid vortexing, vigorous pipetting, and repeated freeze-thaw cycles. Use wide-bore pipette tips and cut pipette tips to widen the opening when transferring viscous DNA solutions. Some protocols recommend gentle mixing by inversion or slow pipetting instead of vortexing.

Common DNA Quality Failures

Library Preparation Optimization

Adapter Ligation and Cleanup

Library preparation converts extracted DNA into a sequencing-ready library by attaching platform-specific adapters. The efficiency of adapter ligation directly affects sequencing yield. If ligation fails, the DNA cannot attach to the flowcell or the pore, and you will observe low read counts.

Check the molar ratio of adapters to DNA. Too little adapter results in incomplete ligation, while too much adapter can cause adapter dimers that consume sequencing capacity without producing useful reads. Follow the manufacturer recommendations for your specific kit and adjust based on your DNA concentration measurement.

Cleanup steps remove unligated adapters, enzymes, and buffer components. Incomplete cleanup can introduce contaminants that block pores or inhibit downstream reactions. Use fresh magnetic beads and verify that the bead-to-sample ratio matches the kit protocol. Ethanol carry-over from bead cleanup can inhibit basecalling and reduce yield.

PCR Amplification Tradeoffs

Some library preparation protocols include a PCR amplification step to increase the amount of library available for sequencing. PCR amplification is necessary when input DNA quantity is low, such as when working with clinical samples or environmental metagenomes. However, PCR introduces amplification bias and can shorten read lengths.

A study of Nanopore adaptive sampling in clinical sputum samples found that PCR amplification necessary to compensate for low DNA extraction yields produced shorter read lengths of approximately 2.5 kilobases. This read length reduction limited the enrichment achieved by adaptive sampling to 3.1-fold instead of the 5 to 10-fold enrichment promised under ideal conditions. The same study encountered rapid pore loss that reduced total sequencing yield by an estimated 80 percent, and the cause could not be determined despite extensive attempts to reduce contaminant carry-over.

This evidence illustrates a practical tradeoff. When you amplify low-yield samples, you gain the ability to sequence at all, but you sacrifice read length and may introduce contaminants that damage the flowcell. For bacterial isolates that can be cultured to high density, avoid PCR amplification by extracting sufficient DNA from the culture. For metagenomic or clinical samples where amplification is unavoidable, plan for reduced performance and consider whether short-read sequencing might be more appropriate for your question.

Library Quantification and Loading

Quantify the final library before loading it onto the flowcell. Underloading the flowcell wastes sequencing capacity and produces low yield. Overloading can cause pore blockage and reduce read quality. Use the quantification method recommended by the platform manufacturer, which may be a fluorometric assay or a quantitative PCR approach.

The optimal loading concentration depends on the flowcell type and the library preparation kit. Start with the manufacturer recommended range and adjust based on your observed pore occupancy. If you see high pore occupancy but low read output, your library may contain contaminants that block pores after initial attachment. If you see low pore occupancy, increase the loading concentration in the next run.

Sequencing Run Conditions

Flowcell Handling and Storage

Flowcells are sensitive to temperature, humidity, and physical damage. Store flowcells at the recommended temperature and allow them to equilibrate to room temperature before loading. Condensation on the flowcell surface can damage the pores or interfere with the sequencing reaction. Inspect the flowcell for visible damage or debris before loading.

Check the expiration date and lot number of each flowcell. Different lots may have different pore densities and performance characteristics. Record the lot number in your run notes so you can identify whether a particular lot is associated with poor performance.

Pore Occupancy and Active Pores

The number of active pores on a flowcell determines the maximum possible sequencing throughput. Monitor pore occupancy during the run using the platform software. A healthy run typically shows high pore occupancy within the first hour and maintains it for the duration of the run.

Rapid pore loss is a common failure mode in Nanopore sequencing. The clinical sputum study described earlier encountered pore loss that reduced total yield by an estimated 80 percent. The researchers ruled out adaptive sampling and pore underloading as contributing factors but could not determine the cause. This example demonstrates that pore loss can occur even when you follow standard protocols and that some failures resist troubleshooting.

If you observe pore loss, check the following variables in your next run. First, verify that your library is clean and free of contaminants that could block pores. Second, confirm that your wash buffer and other reagents are fresh and not expired. Third, check that the flowcell was stored and handled correctly. Fourth, consider whether the sample type itself contains compounds that damage pores, such as polysaccharides or pigments from clinical sputum or environmental samples.

Run Duration and Read Length Over Time

Sequencing runs can last from hours to days depending on the platform, flowcell, and desired throughput. Read length tends to decrease over the course of a run as pores become blocked or the sequencing reaction slows. If you need very long reads for complete genome assembly, you may need to stop the run early and accept lower total yield in exchange for longer reads.

The optimal run duration depends on your assembly goal. For a small bacterial genome of 5 megabases, you may need only a few hours of sequencing to achieve sufficient coverage. For a metagenome or a complex sample with multiple organisms, you may need longer runs to achieve adequate depth for all organisms present.

Basecalling and Data Quality

Choosing a Basecalling Model

Basecalling converts the raw electrical signal from the sequencer into nucleotide sequences. The accuracy of basecalling depends on the model used and the quality of the raw signal. Most platforms offer multiple basecalling models optimized for different read lengths, accuracy levels, and throughput.

For bacterial genome assembly, you generally want the highest accuracy basecalling model available, even if it requires more computational time. The error rate of the raw reads affects the assembly quality and the amount of polishing required. A study benchmarking long-read assembly software on PacBio HiFi metagenomes found that assemblies can include more than 40 errors per 100 million base pairs of assembled contigs, including chimeras, prematurely circularized sequences, haplotyping errors, excessive repeats, and phantom sequences. These errors can be introduced during assembly instead of basecalling, but the quality of the input reads influences how well the assembler performs.

Read Quality Scores and Filtering

After basecalling, examine the read quality scores. Most platforms provide a quality score for each read and for each base within a read. Low-quality reads can be filtered out before assembly to improve assembly quality, but aggressive filtering can remove useful data and reduce coverage.

The appropriate quality threshold depends on your platform and application. For Nanopore data, you may filter reads below a quality score of 7 or 10, depending on the basecalling model and the flowcell chemistry. For PacBio HiFi data, the quality scores are generally higher, and you may filter less aggressively.

Polishing and Error Correction

After assembly, polishing steps use the raw reads to correct errors in the assembled contigs. Polishing can be performed with the same platform data or with short-read data from a separate sequencing run. Hybrid assembly approaches that combine long reads for scaffolding and short reads for error correction can produce highly accurate genomes.

The choice of polishing strategy depends on your accuracy requirements. For a reference-quality genome intended for publication, you may need multiple rounds of polishing. For a draft genome used for gene content analysis, a single polishing round may be sufficient.

Assembly and Post-Processing

Selecting an Assembler

Multiple assemblers are available for long-read bacterial genomes, including Flye, Canu, hifiasm, and others. The choice of assembler affects the contiguity and accuracy of the final assembly. A benchmarking study of four long-read assembly programs on PacBio HiFi metagenomes found that assemblies can include multi-domain chimeras, prematurely circularized sequences, haplotyping errors, excessive repeats, and phantom sequences. These errors are more common in complex metagenomes but can also occur in single-genome assemblies.

For a single bacterial isolate, most assemblers perform well if the input reads are of sufficient quality and coverage. For metagenomes or samples with multiple organisms, the choice of assembler becomes more critical. The benchmarking study provides an open-source tool and reproducible workflow for evaluating assembly errors, which you can use to assess the quality of your own assemblies.

Evaluating Assembly Quality

After assembly, evaluate the quality using multiple metrics. Check the number of contigs, the N50 value, and the total assembly size. Compare the assembly size to the expected genome size for your organism. A complete bacterial genome should assemble into a single circular chromosome, possibly with additional plasmids.

Use assembly evaluation tools to check for misassemblies, chimeras, and other structural errors. The tool described in the benchmarking study can quantify read clipping events, in which long reads are systematically split during mapping to maximize agreement with assembled contigs. These clipping events indicate where assemblies diverge from their source reads.

Completing Circular Genomes

For bacterial genomes, the goal is often to produce a complete circular chromosome. If your assembly produces a linear contig that should be circular, you may need to check for overlapping ends and circularize the sequence manually. Some assemblers perform circularization automatically, while others require manual intervention.

If your assembly contains multiple contigs for a single chromosome, you may need to increase sequencing depth or use a different assembler. Gaps in the assembly often correspond to repetitive regions that could not be resolved with the available read lengths.

Common Failure Patterns and Their Resolution

Low Yield with Adequate Pore Occupancy

If you observe high pore occupancy but low total yield, the problem may be in the library preparation or the basecalling step. Check that your library contains a sufficient concentration of DNA molecules with adapters attached. If the library contains adapter dimers or other small molecules, these may occupy pores without producing useful reads.

Re-quantify your library and run a quality check on a small aliquot before loading the full library. Some platforms provide a library quality check that reports the average read length and the proportion of reads with adapters. Use this information to decide whether to proceed with the run or prepare a new library.

High Error Rates in Assembled Genome

If your assembled genome has a high error rate, the problem may be insufficient sequencing depth, poor basecalling accuracy, or assembly errors. Check the coverage of your assembly and compare it to the recommended coverage for your platform. For Nanopore data, you may need 30 to 50-fold coverage for a complete genome. For PacBio HiFi data, you may need 20 to 30-fold coverage.

If coverage is adequate but errors persist, try a different basecalling model or a different assembler. The benchmarking study found that different assemblers produce different error profiles, so switching assemblers may resolve specific error types.

Chimeric Assemblies

Chimeric assemblies occur when reads from two different organisms or two different regions of the same genome are incorrectly joined. This problem is more common in metagenomes but can occur in single-genome assemblies if the sample is contaminated with another organism.

Check your assembly for chimeras by examining the coverage and GC content along the contig. A sudden change in coverage or GC content may indicate a chimera. Use the assembly evaluation tool from the benchmarking study to identify chimeric contigs and remove them before downstream analysis.

Prematurely Circularized Sequences

Prematurely circularized sequences occur when the assembler joins the ends of a contig incorrectly, producing a circular molecule that does not match the true genome structure. This error can occur in regions with repetitive sequences that confuse the assembler.

Check circularized contigs by comparing the sequence at the join point to the expected sequence. If the join point contains a repeat that is not present in the reference genome, the circularization may be incorrect. Use a different assembler or increase sequencing depth to resolve the issue.

Records and Measurements for Troubleshooting

What to Record for Each Run

Maintain a standardized run log for every sequencing experiment. Include the following information:

• Date and operator name
• Sample identifier and organism
• DNA extraction method and kit lot number
• DNA concentration and purity measurements
• DNA fragment size distribution
• Library preparation kit and lot number
• Adapter ligation efficiency if measured
• Library concentration and loading amount
• Flowcell type, lot number, and expiration date
• Sequencing platform and software version
• Basecalling model and parameters
• Run duration and pore occupancy over time
• Total yield and read length distribution
• Assembly metrics and final genome quality

This record allows you to compare runs and identify the variable that changed when a run failed. Without this record, troubleshooting becomes guesswork.

Establishing a Baseline for Your Platform

Before troubleshooting a failed run, establish a baseline for your platform using a well-characterized control sample. Sequence a reference strain with a known genome sequence and compare your results to the expected values. This baseline tells you what yield, read length, and accuracy you should expect under optimal conditions.

The American Type Culture Collection provides authenticated bacterial strains that can serve as controls. Using an authenticated strain ensures that your control sample is what you think it is, eliminating sample identity as a variable.

When to Escalate to Professional Support

Some failures resist troubleshooting despite systematic effort. The clinical sputum study described earlier encountered pore loss that could not be mitigated despite extensive attempts to reduce contaminant carry-over. In such cases, the appropriate action is to contact the platform manufacturer for technical support.

Escalate to professional support when you have:

• Verified that your DNA quality and library preparation meet the manufacturer specifications
• Repeated the run with fresh reagents and a new flowcell
• Tested multiple basecalling models and assemblers
• Confirmed that the failure is reproducible across multiple samples

Provide your run records to the technical support team. The records allow them to identify issues that may be specific to your platform configuration or reagent lot.

Biosafety and Quality Control Context

Handling Bacterial Samples Safely

Working with bacterial cultures requires adherence to biosafety principles. The Biosafety in Microbiological and Biomedical Laboratories manual from the CDC and NIH provides guidance on risk assessment, containment, decontamination, and laboratory biosafety. Determine the biosafety level appropriate for your organism before beginning work.

Most non-pathogenic laboratory strains such as Escherichia coli K-12 require biosafety level 1 practices. Pathogenic organisms such as Salmonella, Shigella, and Mycobacterium tuberculosis require biosafety level 2 or higher. Clinical samples may contain unknown pathogens and should be handled under biosafety level 2 practices at minimum.

Decontaminate all waste and reusable equipment according to your institutional biosafety protocols. Disinfect work surfaces before and after each session. Use appropriate personal protective equipment, including gloves, lab coats, and eye protection.

Quality Control in Clinical and Food Microbiology

If your sequencing work supports clinical diagnostics or food safety testing, follow the quality standards established by relevant authorities. The FDA Bacteriological Analytical Manual provides official methods for food microbiology that include quality control requirements. The Clinical and Laboratory Standards Institute provides standards for antimicrobial susceptibility testing that may be relevant if your sequencing work supports resistance determination.

The World Health Organization defines antimicrobial resistance as a global public health concern and supports surveillance efforts to track resistant organisms. Long-read sequencing can contribute to this surveillance by providing complete genome sequences that reveal resistance mechanisms and transmission patterns. The NCBI Pathogen Detection database collects and analyzes pathogen sequences for public health surveillance, and you can submit your assembled genomes to this database to contribute to global efforts.

Professional Microbiology Practice

The American Society for Microbiology provides professional education and laboratory practice resources for microbiologists. Staying current with best practices in microbial genomics requires ongoing education, as sequencing technologies and analysis methods evolve rapidly. Participate in professional development opportunities and consult the literature for updates on best practices.

Limitations of Long-Read Sequencing Troubleshooting

Sample-Specific Limitations

Some samples resist improvement regardless of troubleshooting effort. Clinical samples such as sputum contain host cells, mucus, and other contaminants that are difficult to remove completely. Environmental samples may contain humic acids, polysaccharides, and other compounds that inhibit enzymes and block pores.

The clinical sputum study demonstrated that even with extensive troubleshooting, pore loss could not be mitigated. This limitation is inherent to the sample type instead of a failure of the protocol. When working with difficult samples, plan for reduced performance and consider whether alternative approaches such as short-read sequencing or targeted enrichment might be more appropriate.

Platform-Specific Limitations

Each sequencing platform has inherent limitations that cannot be overcome through troubleshooting. Nanopore sequencing has higher raw error rates than PacBio HiFi, though the gap has narrowed with newer chemistry. PacBio requires more input DNA than Nanopore, which may be limiting for low-biomass samples.

The choice of platform should be based on your research question and sample characteristics. For complete bacterial genome assembly from pure cultures, both platforms can produce reference-quality genomes. For metagenomes or low-biomass samples, the platform choice may be constrained by input DNA requirements and error rates.

Cost and Time Considerations

Troubleshooting consumes time and resources. Each failed run represents a cost in reagents, flowcells, and operator time. When a run fails, weigh the cost of troubleshooting against the cost of repeating the experiment with a fresh sample.

For valuable or irreplaceable samples, invest more effort in troubleshooting. For routine samples that can be easily replaced, consider repeating the experiment with a fresh extraction instead of spending extensive time troubleshooting a problematic library.

A Decision Framework for Run Triage and Reagent Lot Tracking

Building a Run-Level Decision Tree

When a sequencing run fails, the natural impulse is to adjust the first variable that comes to mind, often the basecalling model or the assembler. This approach wastes time and reagents because it skips the diagnostic steps that distinguish between a sample problem, a library problem, a flowcell problem, and an analysis problem. A structured decision tree forces you to collect the right data before changing any variable.

Start the triage process by asking whether the failure occurred before, during, or after the sequencing reaction. Failures before the run include low DNA yield, poor purity ratios, and failed adapter ligation. Failures during the run include low pore occupancy, rapid pore loss, and declining read quality over time. Failures after the run include poor basecalling accuracy, low read quality scores, and assembly errors that do not match the quality of the input reads.

For each failure category, define a specific set of measurements that must be collected before you change any protocol parameter. For a pre-run failure, collect the DNA concentration from fluorometry, the absorbance ratios from spectrophotometry, and the fragment size distribution from a TapeStation or pulsed-field gel. For a mid-run failure, collect the pore occupancy trace, the read length distribution over time, and the yield at hourly intervals. For a post-run failure, collect the read quality score distribution, the coverage depth, and the assembly metrics including N50 and number of contigs.

The decision tree should have explicit branches that lead to a repeat run versus a protocol change. If the DNA quality metrics meet the manufacturer specifications and the library quantification is within range, but the run still fails, the problem is likely in the flowcell or the reagents. In this case, repeat the run with a new flowcell and fresh reagents before changing your extraction or library preparation protocol. If the DNA quality metrics are outside the acceptable range, re-extract the DNA before spending a flowcell on a library that cannot produce good data.

A study of Nanopore adaptive sampling in clinical sputum samples illustrates the value of this structured approach. The researchers encountered rapid pore loss that reduced total sequencing yield by an estimated 80 percent. They ruled out adaptive sampling and pore underloading as contributing factors through systematic testing. They could not determine the cause despite extensive attempts to reduce contaminant carry-over. This example shows that a decision tree can eliminate variables even when it cannot identify the root cause. The ability to rule out causes is itself valuable because it tells you which variables not to change in the next run.

Establishing a Reagent Lot Baseline

Reagent lot variation is an underappreciated source of sequencing failures. Extraction kits, library preparation kits, wash buffers, and flowcells can vary between lots in ways that affect performance. If you do not track lot numbers, you cannot identify a bad lot when it causes a run failure.

Create a reagent lot log that records the lot number, expiration date, and date of first use for every kit and buffer that touches your sample. When a run fails, check whether the same lot was used in your most recent successful run. If the lot changed between the successful run and the failed run, the reagent lot is a prime suspect.

The American Type Culture Collection provides authenticated bacterial strains that can serve as control samples for lot testing. When you receive a new lot of an extraction kit or a library preparation kit, test it on a control strain before using it on valuable samples. This practice costs one extra reaction but prevents the loss of an entire sequencing run.

For flowcells, record the lot number and the pore density reported by the manufacturer. Different flowcell lots can have different pore densities and performance characteristics. If you observe consistently low yield across multiple runs with the same flowcell lot, contact the manufacturer with your lot numbers and run records.

The Two-Run Rule for Protocol Changes

When troubleshooting a failed run, change only one variable at a time and test that change on a control sample before applying it to your problem sample. This rule prevents the common error of changing multiple variables simultaneously, which makes it impossible to determine which change fixed the problem.

A practical implementation of this rule is the two-run protocol. After a failed run, identify the single most likely cause based on your run records. Change that one variable and repeat the run with a control sample. If the control sample performs well with the changed variable, apply the same change to your problem sample. If the control sample also fails, the changed variable was not the cause, and you should revert to your original protocol and test a different variable.

This approach requires that you have a control sample available and that you are willing to spend a flowcell on a test run. For laboratories that sequence frequently, this cost is justified because it prevents the waste of multiple flowcells on untested protocol changes. For laboratories that sequence infrequently, the two-run rule may be too expensive, and you should instead contact the platform manufacturer for technical support with your run records in hand.

Distinguishing Sample-Limited from Protocol-Limited Failures

Some samples fail regardless of protocol quality. Clinical samples such as sputum contain host cells, mucus, and other contaminants that are difficult to remove completely. Environmental samples may contain humic acids, polysaccharides, and pigments that inhibit enzymes and block pores. When you have verified that your protocol meets the manufacturer specifications and the failure persists, the limitation may be inherent to the sample.

The clinical sputum study provides a clear example of a sample-limited failure. The researchers achieved at best 3.1-fold enrichment of bacterial sequence output with adaptive sampling due to shorter read lengths from PCR amplification necessary to compensate for low DNA extraction yields. More critically, they encountered rapid pore loss that reduced total sequencing yield by an estimated 80 percent. They could not mitigate the pore loss despite extensive attempts to reduce contaminant carry-over, and they ruled out adaptive sampling and pore underloading as contributing factors.

When you suspect a sample-limited failure, compare the performance of your problem sample to a control sample processed in parallel. If the control sample performs well and the problem sample fails, the issue is in the sample. If both samples fail, the issue is in the protocol, the reagents, or the flowcell.

For sample-limited failures, consider whether an alternative approach might be more appropriate. Short-read sequencing may be more robust to contaminants and degraded DNA. Hybrid approaches that combine short reads for error correction with long reads for scaffolding can produce high-quality assemblies even when the long-read data are limited. The choice of approach depends on your research question and the accuracy requirements of your downstream analysis.

Recording Negative Results

Negative results are as valuable as positive results for troubleshooting. When a protocol change does not fix a failure, record that change and its outcome in your run log. This record prevents you from repeating the same unsuccessful change in the future and helps you identify patterns across multiple failures.

For example, if you try a different basecalling model and the error rate does not improve, record that result. If you try a different assembler and the assembly quality does not improve, record that result. Over time, these negative results build a knowledge base specific to your laboratory, your samples, and your platform configuration.

The benchmarking study of long-read assembly software provides a model for systematic evaluation. The researchers quantified read clipping events, in which long reads are systematically split during mapping to maximize agreement with assembled contigs, to identify where assemblies diverge from their source reads. They found that long-read metagenome assemblies can include more than 40 errors per 100 million base pairs of assembled contigs, including multi-domain chimeras, prematurely circularized sequences, haplotyping errors, excessive repeats, and phantom sequences. They provided an open-source tool and a reproducible workflow for rigorous evaluation of assembly errors.

You can apply the same principle to your troubleshooting. Instead of relying on anecdotal impressions of whether a change helped, use quantitative metrics to evaluate the outcome. Record the yield, read length distribution, quality scores, and assembly metrics before and after each protocol change. This quantitative approach turns troubleshooting from guesswork into a systematic optimization process.

Escalation Criteria for Professional Support

Professional support from the platform manufacturer should be your escalation path when you have exhausted your own troubleshooting capacity. The decision to escalate should be based on objective criteria instead of frustration or time pressure.

Escalate to professional support when you have verified that your DNA quality and library preparation meet the manufacturer specifications, repeated the run with fresh reagents and a new flowcell, tested multiple basecalling models and assemblers, and confirmed that the failure is reproducible across multiple samples. Provide your complete run records to the technical support team, including DNA quality metrics, library preparation details, flowcell lot numbers, pore occupancy data, and assembly metrics.

The Biosafety in Microbiological and Biomedical Laboratories manual from the CDC and NIH provides guidance on risk assessment and containment that is relevant when working with pathogenic organisms. If your sequencing work involves pathogens, ensure that your laboratory practices meet the appropriate biosafety level before escalating technical issues. The American Society for Microbiology provides professional education resources that can help you stay current with best practices in microbial genomics.

For laboratories that support clinical diagnostics or food safety testing, the FDA Bacteriological Analytical Manual and the Clinical and Laboratory Standards Institute provide quality standards that may apply to your sequencing workflow. The World Health Organization defines antimicrobial resistance as a global public health concern and supports surveillance efforts that can benefit from long-read sequencing data. The NCBI Pathogen Detection database collects and analyzes pathogen sequences for public health surveillance, and you can submit your assembled genomes to this database to contribute to global efforts.

Implementing the Framework in Your Laboratory

Implementing this decision framework requires a commitment to record keeping and a willingness to spend resources on control samples and test runs. Start by creating a standardized run log template that includes all the fields described in the Records and Measurements section. Train all laboratory members who operate the sequencer to use the template consistently.

Next, establish a baseline for your platform using a control strain from the American Type Culture Collection. Sequence this control strain under your standard protocol and record the expected yield, read length, and accuracy. This baseline gives you a reference point for evaluating future runs.

Finally, adopt the two-run rule for protocol changes and the escalation criteria for professional support. These practices will not prevent all failures, but they will ensure that you learn from each failure and avoid repeating the same mistakes. The clinical sputum study demonstrates that some failures resist troubleshooting despite systematic effort. When you encounter such a failure, the appropriate response is to document what you have ruled out, escalate to professional support, and consider whether an alternative sequencing approach might better suit your research question.

Frequently Asked Questions

Why did my Nanopore run produce very few reads?

Low read output typically results from insufficient library loading, degraded DNA, or contaminants that block pores. Check your library concentration with a fluorometer and verify that the DNA fragment size distribution meets the platform requirements. If the library looks acceptable, check the pore occupancy during the run to determine whether pores were active but unproductive or blocked from the start.

What is the minimum DNA quality needed for long-read sequencing?

The minimum DNA quality depends on the platform and library preparation kit. Most manufacturers specify a minimum concentration and fragment length. As a general practice, aim for a 260/280 ratio near 1.8 and a 260/230 ratio above 2.0. Use fluorometry to confirm that the majority of your nucleic acid is double-stranded DNA instead of RNA or single-stranded DNA.

How can I prevent pore loss during a Nanopore run?

Pore loss can result from contaminants in the library, expired reagents, or flowcell damage. Purify your library thoroughly, use fresh wash buffers, and handle flowcells according to the manufacturer instructions. Note that some samples cause pore loss that cannot be prevented, as demonstrated in a study of clinical sputum samples where pore loss reduced yield by an estimated 80 percent despite extensive troubleshooting.

Should I use PCR amplification in my library preparation?

PCR amplification is necessary when input DNA is limiting, but it shortens read lengths and can introduce bias. A study of Nanopore adaptive sampling found that PCR amplification necessary for low-yield clinical samples produced read lengths of approximately 2.5 kilobases, which limited the enrichment achieved. For bacterial isolates that can be cultured, extract sufficient DNA to avoid PCR amplification.

Why does my assembled genome contain errors despite high coverage?

Assembly errors can result from repetitive regions, haplotyping errors, or assembler-specific artifacts. A benchmarking study of long-read assembly software found that assemblies can include more than 40 errors per 100 million base pairs, including chimeras and prematurely circularized sequences. Try a different assembler, increase sequencing depth, or use the evaluation tool from the benchmarking study to identify and correct specific error types.

How do I know if my assembly is complete?

A complete bacterial genome assembly should produce a single circular chromosome, possibly with additional circular plasmids. Compare the total assembly size to the expected genome size for your organism. Check that all expected genes are present and that the assembly has no gaps. Use assembly evaluation tools to check for structural errors such as chimeras or misjoins.

What should I do if my sequencing run fails repeatedly?

If you have verified DNA quality, library preparation, and run conditions, and the failure persists, contact the platform manufacturer for technical support. Provide your run records, including DNA quality metrics, library preparation details, and pore occupancy data. The manufacturer may identify issues with your specific platform configuration or reagent lot.

Can I use long-read sequencing for metagenomic samples?

Long-read sequencing can be used for metagenomes, but performance is often limited by sample characteristics. A study of clinical sputum metagenomes found that adaptive sampling achieved only 3.1-fold enrichment instead of the promised 5 to 10-fold, and pore loss reduced yield by an estimated 80 percent. For complex metagenomes, consider whether short-read sequencing or a hybrid approach might better suit your research question.

Related Bacteriology Guides

• Bacterial Pathogens in Poultry: Salmonella, Escherichia coli, and Other Common Agents
• Avian Bacterial Infections in Poultry: Comprehensive Review of Common Pathogens, Clinical Signs, and Diagnostic Approaches
• Which Bacteria Are Common to Raw Poultry? A Safety and Pathogen Guide
• Bacteria on Chicken: Common Pathogens and Mitigation in Poultry Production
• Common Bacterial Pathogens in Chickens: Identification and Antimicrobial Management

References and Further Reading

• Biosafety in Microbiological and Biomedical Laboratories. CDC and NIH.
• Microbiology Resources. American Society for Microbiology.
• Antimicrobial Resistance. World Health Organization.
• Bacteriological Analytical Manual. U.S. Food and Drug Administration.
• Antimicrobial Susceptibility Testing. Clinical and Laboratory Standards Institute.
• Culture Collections. American Type Culture Collection.
• NCBI Pathogen Detection. National Center for Biotechnology Information.
• Troubleshooting common errors in assemblies of long-read metagenomes.. Nature biotechnology, 2026.
• Probing the limits of genetic recoding using multi-omics-guided evolution.. 2026.
• Limited value of Nanopore adaptive sampling in a long-read metagenomic profiling workflow of clinical sputum samples.. 2025.
• De-novo assembly of 82 bacterial genomes using Nanopore sequencing and prediction of biosynthetic capacity.. 2026.

This article is educational and does not replace validated laboratory procedures, clinical diagnosis, institutional biosafety review, or public-health guidance.