A Step-by-Step Guide to Bacterial Genome Assembly: From Raw Reads to Contigs

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

Quick Answer

• Assemble bacterial genomes by assessing raw read quality, trimming adapters and low-quality bases, correcting sequencing errors, selecting an assembler such as SPAdes, and evaluating contig quality before annotation.
• Start with quality assessment of raw reads, then choose assembly parameters based on read length and sequencing platform, with SPAdes as the default for short-read isolate data.
• Assembly quality depends heavily on input data cleanliness, and contaminated or low-quality reads will produce fragmented or misleading contigs regardless of assembler choice.

At a Glance

Understanding Bacterial Genome Assembly

Bacterial whole genome sequencing has become a routine laboratory procedure because modern DNA sequencing instruments generate very large amounts of data at relatively low cost. The subsequent step of assembling those reads into contiguous sequences, called contigs, is well addressed by existing software tools. The final step of annotating all relevant genomic features on those contigs can be achieved slowly using existing web and email based systems, but these are not applicable for sensitive data or for integrating into computational pipelines. A command line software tool called Prokka can fully annotate a draft bacterial genome in about 10 minutes on a typical desktop computer, producing standards compliant output files for further analysis or viewing in genome browsers [<a href="#ref-1">1</a>].

The assembly process transforms millions of short DNA fragments into a representation of the bacterial chromosome and any plasmids present in the sample. This representation is rarely a single complete circular chromosome. Instead, most assemblies produce a set of contigs that must be ordered and oriented relative to one another. The quality of this final representation determines what biological questions can be answered from the data. A fragmented assembly with hundreds of contigs may still be useful for identifying genes and detecting antimicrobial resistance determinants, but it will not support fine-scale comparisons such as identifying subtle intraspecies variation or developing sensitive species-specific detection tools [<a href="#ref-2">2</a>].

Acquisition of high quality bacterial genomes is fundamental when the goal is investigating subtle intraspecies variation or developing sensitive species-specific tools for detection and identification of pathogens. Pacific Biosciences technology is attractive for this purpose because it generates reads longer than 10,000 base pairs. A bacterial genome assembly pipeline based on open source software can be handled by non-bioinformaticians interested in transforming sequencing data into reliable biological information. Using this method, six Dickeya solani genomes were successfully closed, and the assembly process ran on a slightly improved desktop computer [<a href="#ref-2">2</a>].

Whole genome assembly of bacterial genomes with short reads is a common task because DNA sequencing has become cheaper with advances in technology. The process of assembling a genome has no absolute golden standard, and it requires performing a sequence of steps, each of which can involve combinations of many different tools. The quality of the final assembly is always strongly related to the quality of the input data. A package called WGA-LP connects state-of-the-art programs for microbial analysis with novel scripts to check and improve the quality of both samples and resulting assemblies. WGA-LP, with its conservative decontamination approach, has shown the capability to create high quality assemblies even in the case of contaminated reads [<a href="#ref-3">3</a>].

Why Assembly Quality Matters for Downstream Analysis

The biological conclusions drawn from a bacterial genome sequence depend entirely on the quality of the assembly. If the assembly contains errors, misjoins, or missing regions, then every downstream analysis built on that assembly inherits those problems. Gene annotation will miss or misidentify coding sequences. Antimicrobial resistance gene detection may produce false negatives if resistance determinants fall in unassembled regions. Phylogenetic comparisons may place a strain in the wrong clade if assembly errors create spurious single nucleotide polymorphisms.

For public health applications such as pathogen genomics and genomic surveillance, assembly quality directly affects outbreak investigations. A fragmented assembly may still allow species identification and basic typing, but it may not support the fine-grained comparisons needed to determine whether two clinical isolates are part of the same transmission chain. The same limitation applies to antimicrobial resistance surveillance, where the presence or absence of specific resistance genes must be determined with confidence. The World Health Organization recognizes antimicrobial resistance as a global public health concern and emphasizes surveillance for tracking resistance trends [<a href="#ref-4">4</a>].

The Relationship Between Input Data Quality and Assembly Output

The quality of the final assembly is strongly related to the quality of the input data. This relationship holds regardless of which assembler is selected. Low-quality reads with high error rates create ambiguous overlaps that assemblers must resolve, often incorrectly. Contaminated reads from other organisms introduce foreign sequences that either assemble into spurious contigs or interfere with the assembly of the target genome. Adapter sequences that remain in the reads create artificial overlaps that mislead the assembler into joining regions that are not actually adjacent in the genome [<a href="#ref-3">3</a>].

The practical implication is that quality control must happen before assembly, not after. Researchers should assess raw read quality, trim adapters and low-quality bases, and check for contamination before running an assembler. This upfront investment of time reduces downstream troubleshooting and produces assemblies that are more likely to be biologically accurate.

Sequencing Platforms and Their Impact on Assembly Strategy

The choice of sequencing platform determines the assembly strategy because read length and error profile differ substantially between technologies. Short-read platforms such as Illumina produce reads of 150 to 300 base pairs with very low error rates. Long-read platforms such as Pacific Biosciences and Oxford Nanopore produce reads of thousands to tens of thousands of base pairs but with higher error rates that require different correction strategies.

Short-Read Assembly with Illumina Data

Illumina sequencing remains the most common platform for bacterial whole genome sequencing because of its low cost per base and high accuracy. SPAdes was originally developed for de novo assembly of genome sequencing data produced for cultivated microbial isolates and for single-cell genomic DNA sequencing. The functionality of SPAdes was later extended to enable assembly of IonTorrent data, as well as hybrid assembly from short and long reads from PacBio and Oxford Nanopore platforms [<a href="#ref-5">5</a>].

For short-read data, the assembly process begins with quality trimming to remove low-quality bases and adapter sequences. The trimmed reads are then subjected to error correction, which identifies and corrects sequencing errors before the assembly step. SPAdes includes built-in error correction that works well for most bacterial datasets. The assembler then builds a de Bruijn graph from the corrected reads and resolves the graph to produce contigs [<a href="#ref-5">5</a>].

The choice of k-mer size affects assembly quality. Smaller k-mers capture more of the genome but create more ambiguous paths in the graph. Larger k-mers produce more specific overlaps but may miss regions with lower coverage. SPAdes automatically selects k-mer sizes based on the read length, but users can override these defaults when working with unusual datasets [<a href="#ref-5">5</a>].

Long-Read Assembly with PacBio and Oxford Nanopore Data

Long-read sequencing platforms generate reads that span repetitive regions and often cover entire genes or operons. This long-range information simplifies the assembly problem because the assembler does not need to resolve as many ambiguous overlaps. The tradeoff is that long-read platforms have higher error rates than short-read platforms, so error correction becomes a critical step [<a href="#ref-2">2</a>].

Pacific Biosciences technology generates reads longer than 10,000 base pairs, which makes it highly attractive for generating closed bacterial genomes. A pipeline based on open source software can be handled by non-bioinformaticians, and the assembly process can run on a slightly improved desktop computer. This accessibility makes long-read assembly feasible for laboratories without dedicated bioinformatics infrastructure [<a href="#ref-2">2</a>].

Long-read assembly typically proceeds through three stages. First, the raw reads are corrected using either the sequencing platform's own consensus algorithm or a standalone correction tool. Second, the corrected reads are assembled using a long-read assembler such as Canu, Flye, or the long-read mode of SPAdes. Third, the resulting assembly is polished using the original raw reads to correct any remaining errors.

Hybrid Assembly Approaches

Hybrid assembly combines short reads and long reads from the same sample to leverage the strengths of both platforms. The long reads provide long-range context that resolves repeats and ordering, while the short reads provide high accuracy that corrects long-read errors. SPAdes includes a hybrid assembly mode that accepts both short and long reads as input [<a href="#ref-5">5</a>].

Hybrid assembly is particularly useful for generating closed genomes when long-read coverage is insufficient for a complete long-read-only assembly. The short reads fill in gaps and correct errors, while the long reads provide the scaffolding information needed to order contigs into a single chromosome. This approach is more complex than single-platform assembly because it requires coordinating two sequencing runs and managing two datasets, but it often produces the highest quality assemblies.

Step-by-Step Assembly Workflow

The assembly workflow proceeds through a series of discrete stages, each with specific inputs, outputs, and quality checks. The following protocol describes the process for assembling an isolate bacterial dataset using SPAdes, with notes on how to adapt the workflow for other platforms and data types [<a href="#ref-5">5</a>].

Step 1: Assess Raw Read Quality

Before any processing, examine the raw sequencing data to understand its quality characteristics. Run quality assessment tools on the raw FASTQ files to generate per-base quality scores, GC content distributions, adapter content, and duplication levels. These metrics reveal whether the sequencing run was successful and whether the data are suitable for assembly.

Key observations to record at this stage include the total number of reads, the read length distribution, the mean quality score, and the GC content. Bacterial genomes typically have GC content between 25 and 75 percent depending on the species. A GC distribution that differs dramatically from the expected value for the target organism may indicate contamination or sequencing bias.

Step 2: Trim and Filter Reads

Trim adapter sequences and low-quality bases from the raw reads. Adapter trimming removes the sequencing adapters that remain at the ends of reads when the insert size is shorter than the read length. Quality trimming removes bases with quality scores below a threshold, typically at the 3-prime end of the read where quality tends to decline.

Filtering removes entire reads that fail quality thresholds. Reads that are too short after trimming, reads with very low mean quality, and reads that contain ambiguous bases may all be removed. The filtering thresholds should be conservative to avoid discarding useful data. Overly aggressive filtering reduces coverage and can fragment the assembly.

Record the number of reads retained after trimming and filtering. A large drop in read count may indicate a problem with the sequencing run or with the trimming parameters.

Step 3: Check for Contamination

Contamination is a common cause of poor assembly quality. Contaminating sequences may come from the laboratory environment, from reagents, from other samples processed on the same sequencing run, or from the host organism if the sample was clinical or environmental. Contamination introduces foreign reads that either assemble into spurious contigs or interfere with the assembly of the target genome [<a href="#ref-3">3</a>].

Several approaches can detect contamination. The GC content distribution may show a secondary peak if a contaminant with different GC content is present. Taxonomic classification of reads using a tool such as Kraken or Kaiju can identify reads from unexpected organisms. Coverage analysis after assembly can reveal contigs with markedly different coverage than the rest of the genome, which often indicates contamination.

WGA-LP uses a conservative decontamination approach that has shown the capability to create high quality assemblies even in the case of contaminated reads. This package connects state-of-the-art programs for microbial analysis with novel scripts to check and improve the quality of both samples and resulting assemblies [<a href="#ref-3">3</a>].

Step 4: Correct Sequencing Errors

Error correction identifies and corrects sequencing errors before assembly. This step is essential for long-read data, where error rates are high, and beneficial for short-read data, where errors are less frequent but still present. SPAdes includes built-in error correction that works well for most bacterial datasets [<a href="#ref-5">5</a>].

The error correction step uses the redundancy of sequencing coverage to distinguish true sequence variants from sequencing errors. Positions where the majority of reads agree are considered correct, while positions where a minority of reads disagree are considered errors and corrected to match the majority. This approach works well when coverage is sufficient, typically above 20-fold for short-read data.

Step 5: Select Assembly Parameters

Choose the assembler and parameters based on the sequencing platform and the biological question. For short-read isolate data, SPAdes is a reliable default choice. For long-read data, consider Canu, Flye, or the long-read mode of SPAdes. For hybrid data, use the hybrid mode of SPAdes or a dedicated hybrid assembler [<a href="#ref-5">5</a>].

The k-mer size is the most important parameter for de Bruijn graph assemblers such as SPAdes. Smaller k-mers produce more sensitive assemblies that capture more of the genome but create more ambiguous paths. Larger k-mers produce more specific assemblies with fewer ambiguities but may miss regions with lower coverage. SPAdes automatically selects k-mer sizes based on read length, but users can override these defaults [<a href="#ref-5">5</a>].

Step 6: Run the Assembly

Run the assembler with the selected parameters. For SPAdes, the basic command for isolate data is straightforward and requires only the input reads and an output directory. The assembler produces a set of output files, including the assembled contigs in FASTA format, assembly statistics, and intermediate files that can be useful for troubleshooting [<a href="#ref-5">5</a>].

The assembly run time depends on the genome size, coverage, read length, and available computational resources. A typical bacterial genome with 50-fold coverage of 150-base-pair reads assembles in minutes to hours on a modern desktop computer. Long-read assemblies may take longer because of the error correction step.

Step 7: Evaluate Assembly Quality

Evaluate the assembly using both global statistics and targeted checks. Global statistics include the number of contigs, the total assembly length, the N50 value, and the largest contig. The N50 is the contig length at which half of the assembled bases are in contigs of that length or longer. Higher N50 values indicate more contiguous assemblies.

Compare the assembly statistics against the expected genome size for the target species. A total assembly length that is much larger than expected may indicate contamination. A total assembly length that is much smaller than expected may indicate that the assembly is incomplete or that the sample contained multiple strains.

Completeness and contamination can be estimated using tools such as CheckM, which uses lineage-specific marker genes to estimate the proportion of the genome that is present and the proportion that comes from contaminating organisms. These estimates provide a more rigorous assessment than simple statistics.

Step 8: Curate the Assembly

Curation involves removing contaminating contigs, splitting misassembled regions, and optionally ordering and orienting contigs relative to a reference genome. Contaminating contigs can be identified by coverage analysis, taxonomic classification, or comparison against reference databases. Misassembled regions can be identified by examining coverage patterns and by comparing the assembly against a closely related reference genome.

For many applications, curation is optional. A draft assembly with some contamination may still be suitable for gene detection and antimicrobial resistance screening. For applications that require a closed genome, such as fine-scale phylogenetic analysis, curation is essential.

Step 9: Annotate the Assembly

Annotation identifies the genomic features on the assembled contigs, including coding sequences, tRNA genes, rRNA genes, and other functional elements. Prokka is a command line software tool that can fully annotate a draft bacterial genome in about 10 minutes on a typical desktop computer. It produces standards compliant output files for further analysis or viewing in genome browsers [<a href="#ref-1">1</a>].

The annotation step is the final stage of the assembly pipeline and produces the biologically interpretable output. The annotated genome can then be used for downstream analyses such as antimicrobial resistance gene detection, multilocus sequence typing, phylogenetic analysis, and comparative genomics.

SPAdes in Detail

SPAdes, the St. Petersburg genome Assembler, was originally developed for de novo assembly of genome sequencing data produced for cultivated microbial isolates and for single-cell genomic DNA sequencing. With time, the functionality of SPAdes was extended to enable assembly of IonTorrent data, as well as hybrid assembly from short and long reads from PacBio and Oxford Nanopore platforms [<a href="#ref-5">5</a>].

SPAdes Assembly Pipelines

The SPAdes package includes protocols for five different assembly pipelines. The first is for assembling isolate bacterial datasets, which is the most common use case. The second is for assembling metagenomic datasets, where the input contains reads from multiple organisms. The third is for assembling sets of putative plasmids, which requires different parameters because plasmids are often circular and may be present at different copy numbers than the chromosome. The fourth is for assembling transcriptomes, which requires strand-specificity information. The fifth is for assembling putative biosynthetic gene clusters from whole-genome sequencing and metagenomic datasets [<a href="#ref-5">5</a>].

Each pipeline uses different parameters and may require different input formats. The isolate pipeline is the simplest and is appropriate for most bacterial genome assembly projects. The metagenomic pipeline is more complex and requires careful attention to parameter selection to avoid chimeric assemblies.

Installing and Running SPAdes

SPAdes is freely available under an open source license and can be installed on Linux, macOS, and Windows systems. The installation process is straightforward and typically involves downloading the software and adding it to the system path. SPAdes can be run from the command line or through a YAML configuration file that specifies the input data and parameters [<a href="#ref-5">5</a>].

The command line interface accepts the input reads and an output directory as the primary arguments. Additional parameters control the k-mer sizes, the coverage cutoff, and the assembly mode. The YAML format provides a more structured way to specify input data, which is useful for complex datasets with multiple read files or paired-end data [<a href="#ref-5">5</a>].

SPAdes supports restarting a previous run, which is useful when a run fails partway through or when the user wants to adjust parameters without starting from scratch. The restart functionality preserves the intermediate files and resumes from the last completed stage [<a href="#ref-5">5</a>].

Interpreting SPAdes Output

SPAdes produces several output files in the specified output directory. The primary output is the assembled contigs in FASTA format. The assembly statistics are reported in a separate file and include the number of contigs, the total length, the N50, and other metrics. The intermediate files include the error-corrected reads, the assembly graph, and various other files that can be useful for troubleshooting [<a href="#ref-5">5</a>].

The contigs file is the input for downstream analysis. The contigs are sorted by length, with the longest contig first. The header of each contig includes the contig identifier and length information. The assembly graph file can be visualized using tools such as Bandage to understand the structure of the assembly and to identify regions that may require manual curation.

Long-Read Assembly with PacBio Data

Pacific Biosciences technology generates reads longer than 10,000 base pairs, which makes it highly attractive for generating closed bacterial genomes. The long read length allows the assembler to span repetitive regions and to resolve the circular structure of the bacterial chromosome without the ambiguity that plagues short-read assemblies [<a href="#ref-2">2</a>].

PacBio Assembly Pipeline

A bacterial genome assembly pipeline based on open source software can be handled by non-bioinformaticians interested in transforming sequencing data into reliable biological information. The pipeline begins with quality assessment of the raw reads, followed by error correction, assembly, and polishing [<a href="#ref-2">2</a>].

The error correction step is critical for PacBio data because the raw reads have error rates of approximately 10 to 15 percent. The correction step uses the coverage redundancy to identify and correct errors, producing reads with error rates below 1 percent. The corrected reads are then assembled using a long-read assembler.

The assembly process for PacBio data can run on a slightly improved desktop computer, making it accessible to laboratories without dedicated computational infrastructure. Using this method, six Dickeya solani genomes were successfully closed, demonstrating the feasibility of the approach for generating complete bacterial genomes [<a href="#ref-2">2</a>].

Advantages and Limitations of Long-Read Assembly

The primary advantage of long-read assembly is the ability to generate closed genomes. A closed genome is a single circular sequence that represents the complete chromosome. Closed genomes are essential for fine-scale analyses such as identifying subtle intraspecies variation and developing sensitive species-specific detection tools [<a href="#ref-2">2</a>].

The primary limitation of long-read assembly is the higher cost per base compared to short-read sequencing. Long-read platforms also have higher error rates, which require additional computational steps to correct. The combination of higher cost and additional computational requirements makes long-read assembly less accessible than short-read assembly for laboratories with limited resources.

Handling Contaminated Reads

Contamination is one of the most common causes of poor assembly quality. The quality of the final assembly is always strongly related to the quality of the input data. Contaminated reads introduce foreign sequences that either assemble into spurious contigs or interfere with the assembly of the target genome [<a href="#ref-3">3</a>].

Sources of Contamination

Contamination can enter a sequencing project at multiple points. Laboratory contamination can introduce DNA from the environment, from reagents, or from other samples processed in the same laboratory. Index hopping or barcode misassignment can introduce reads from other samples on the same sequencing run. Host contamination can introduce reads from the host organism when the sample is clinical or environmental.

The risk of contamination is higher for samples with low biomass, where the ratio of contaminating DNA to target DNA is higher. Samples such as clinical specimens, environmental samples, and samples from culture collections may all contain contaminating DNA.

Detecting Contamination

Several approaches can detect contamination in sequencing data. The GC content distribution may show a secondary peak if a contaminant with different GC content is present. Taxonomic classification of reads can identify reads from unexpected organisms. Coverage analysis after assembly can reveal contigs with markedly different coverage than the rest of the genome.

WGA-LP uses a conservative decontamination approach that has shown the capability to create high quality assemblies even in the case of contaminated reads. The package connects state-of-the-art programs for microbial analysis with novel scripts to check and improve the quality of both samples and resulting assemblies [<a href="#ref-3">3</a>].

Removing Contamination

Once contamination is detected, the contaminated reads or contigs must be removed before the assembly is used for downstream analysis. Read-level decontamination removes reads that are classified as coming from non-target organisms. Contig-level decontamination removes contigs that are identified as contaminants after assembly.

The decontamination approach should be conservative to avoid removing legitimate sequences. Overly aggressive decontamination can remove plasmids, phages, or other mobile genetic elements that are part of the target genome. The conservative approach used by WGA-LP prioritizes retaining legitimate sequences while removing clear contaminants [<a href="#ref-3">3</a>].

Quality Assessment and Assembly Evaluation

Assembly evaluation is a critical step that determines whether the assembly is suitable for downstream analysis. The evaluation should include both global statistics and targeted checks for completeness and contamination.

Assembly Statistics

The most commonly reported assembly statistics are the number of contigs, the total assembly length, the N50, and the largest contig. The N50 is the contig length at which half of the assembled bases are in contigs of that length or longer. Higher N50 values indicate more contiguous assemblies.

The number of contigs and the N50 are inversely related. A more contiguous assembly has fewer contigs and a higher N50. The total assembly length should be close to the expected genome size for the target species. A total length that is much larger than expected may indicate contamination, while a total length that is much smaller may indicate an incomplete assembly.

Completeness and Contamination Estimates

Completeness and contamination can be estimated using tools such as CheckM, which uses lineage-specific marker genes to estimate the proportion of the genome that is present and the proportion that comes from contaminating organisms. These estimates provide a more rigorous assessment than simple statistics.

A high-quality assembly should have completeness above 95 percent and contamination below 5 percent. Assemblies that fall below these thresholds may still be useful for some applications but should be interpreted with caution.

Comparing Against Reference Genomes

Comparing the assembly against a closely related reference genome can identify misassemblies and structural errors. The comparison can be performed using tools such as QUAST, which aligns the assembly against a reference and reports statistics such as the number of misassemblies, the number of mismatches, and the number of indels.

The comparison against a reference genome is particularly useful for identifying misjoins, where the assembler incorrectly joins two regions that are not adjacent in the genome. Misjoins can create chimeric contigs that produce incorrect biological conclusions.

Common Failure Patterns and Troubleshooting

Several failure patterns recur in bacterial genome assembly projects. Recognizing these patterns and knowing how to address them can save substantial time and frustration.

Fragmented Assembly

A fragmented assembly has many contigs and a low N50. This pattern often indicates that the input data quality was poor, that coverage was insufficient, or that the k-mer size was inappropriate. Check the raw read quality, increase coverage if possible, and adjust the k-mer size.

Assembly Larger Than Expected Genome Size

An assembly that is much larger than the expected genome size often indicates contamination. Check the GC content distribution, perform taxonomic classification of reads, and examine coverage patterns to identify contaminating contigs.

Assembly Smaller Than Expected Genome Size

An assembly that is much smaller than the expected genome size may indicate that the sample contained multiple strains, that coverage was insufficient, or that the assembly parameters were too aggressive. Check the read quality, examine the coverage distribution, and adjust the assembly parameters.

High Error Rate in the Assembly

A high error rate in the assembly may indicate that error correction was insufficient or that the polishing step was skipped. Run additional error correction or polishing steps to improve the assembly accuracy.

Chimeric Contigs

Chimeric contigs are contigs that contain sequences from two different organisms or from two non-adjacent regions of the same genome. Chimeric contigs can be identified by examining coverage patterns, by comparing against reference genomes, or by taxonomic classification of contigs.

Records and Documentation

Maintaining detailed records of the assembly process is essential for reproducibility and for troubleshooting. The records should include the version of each software tool, the parameters used, the input data files, and the output files.

What to Record

Record the following information for each assembly project:

Why Records Matter

Detailed records allow other researchers to reproduce the assembly and to understand the decisions that were made. Records also support troubleshooting when problems are identified later. If a downstream analysis produces unexpected results, the assembly records can help determine whether the problem lies in the assembly or in the downstream analysis.

Biosafety and Laboratory Practice

Bacterial genome assembly is a computational process, but the sequencing data originate from laboratory work that must follow appropriate biosafety practices. The Biosafety in Microbiological and Biomedical Laboratories provides principles for risk assessment, containment, decontamination, and laboratory biosafety [<a href="#ref-6">6</a>]. These principles apply to the handling of bacterial cultures and to the preparation of sequencing libraries.

Risk Assessment

Before working with a bacterial isolate, conduct a risk assessment to determine the biosafety level required for the organism. The risk assessment should consider the pathogenicity of the organism, the route of transmission, and the availability of treatment. The Biosafety in Microbiological and Biomedical Laboratories provides guidance on risk assessment and on the selection of appropriate containment levels [<a href="#ref-6">6</a>].

Containment and Decontamination

Work with bacterial cultures should be conducted at the appropriate biosafety level, using the containment equipment and practices specified for that level. Decontamination procedures should be in place for work surfaces, equipment, and waste. The Biosafety in Microbiological and Biomedical Laboratories provides guidance on decontamination procedures [<a href="#ref-6">6</a>].

Professional Practice

The American Society for Microbiology provides professional microbiology education and laboratory practice context [<a href="#ref-7">7</a>]. Following established professional standards for laboratory practice, including proper documentation, quality control, and ethical conduct, is essential for producing reliable sequencing data.

Antimicrobial Resistance Context

Bacterial genome assembly is a foundational step for antimicrobial resistance surveillance. The World Health Organization defines antimicrobial resistance as a global public health concern and emphasizes the importance of surveillance for tracking resistance trends [<a href="#ref-4">4</a>]. Genome sequencing provides a powerful tool for antimicrobial resistance surveillance because it can detect resistance genes and predict resistance phenotypes.

Genomic Surveillance for Antimicrobial Resistance

Genomic surveillance for antimicrobial resistance involves sequencing bacterial isolates and analyzing their genomes for resistance determinants. The assembly is the first step in this process, followed by gene detection and resistance prediction. The quality of the assembly directly affects the reliability of resistance gene detection.

The Clinical and Laboratory Standards Institute provides standards for antimicrobial susceptibility testing, which is the phenotypic counterpart to genomic resistance prediction [<a href="#ref-8">8</a>]. Genomic prediction of resistance must be validated against phenotypic testing to ensure accuracy.

Public Health Applications

The National Center for Biotechnology Information maintains the NCBI Pathogen Detection system, which collects and analyzes bacterial genomes for public health surveillance [<a href="#ref-9">9</a>]. This system depends on high-quality assemblies from submitted isolates. The quality of the assembly determines whether the isolate can be included in the analysis and whether the results are reliable.

Professional Escalation Criteria

Some assembly problems require escalation to a bioinformatics specialist or to the sequencing facility. Escalate when:

• The assembly fails to complete or crashes repeatedly
• The assembly statistics are far outside the expected range for the target organism
• The completeness and contamination estimates indicate a serious problem
• The assembly contains many chimeric contigs that cannot be resolved with standard approaches
• The downstream analysis produces results that conflict with phenotypic observations

When escalating, provide the assembly records, the raw data files, and a description of the problem. This information allows the specialist to diagnose the issue and to recommend appropriate corrective actions.

Frequently Asked Questions

What is the minimum coverage needed for bacterial genome assembly?

The minimum coverage depends on the sequencing platform and the assembly goals. For short-read Illumina data, coverage of at least 30-fold is generally recommended for a draft assembly, with higher coverage improving the chances of a complete assembly. For long-read data, lower coverage may be sufficient because the long reads provide more information per read. The quality of the final assembly is always strongly related to the quality of the input data, so higher coverage with good quality reads produces better assemblies [<a href="#ref-3">3</a>].

How do I choose between short-read and long-read sequencing for bacterial assembly?

The choice depends on the biological question and the available resources. Short-read sequencing is less expensive and more accessible, making it suitable for most applications including gene detection and antimicrobial resistance screening. Long-read sequencing generates reads longer than 10,000 base pairs, which enables the generation of closed genomes that are essential for fine-scale analyses such as identifying subtle intraspecies variation [<a href="#ref-2">2</a>]. Long-read assembly can run on a slightly improved desktop computer, making it accessible to non-bioinformaticians [<a href="#ref-2">2</a>].

What k-mer size should I use for SPAdes assembly?

SPAdes automatically selects k-mer sizes based on the read length, and the automatic selection works well for most bacterial datasets [<a href="#ref-5">5</a>]. Users can override the automatic selection when working with unusual datasets, such as those with very low coverage or those from organisms with extreme GC content. Smaller k-mers capture more of the genome but create more ambiguous paths, while larger k-mers produce more specific assemblies but may miss regions with lower coverage.

How can I tell if my assembly is contaminated?

Several indicators suggest contamination. The GC content distribution may show a secondary peak if a contaminant with different GC content is present. The total assembly length may be much larger than the expected genome size. Coverage analysis may reveal contigs with markedly different coverage than the rest of the genome. Taxonomic classification of reads or contigs can identify sequences from unexpected organisms. WGA-LP uses a conservative decontamination approach that can create high quality assemblies even in the case of contaminated reads [<a href="#ref-3">3</a>].

What is the N50 and why does it matter?

The N50 is the contig length at which half of the assembled bases are in contigs of that length or longer. A higher N50 indicates a more contiguous assembly with fewer and longer contigs. The N50 is a useful summary statistic for comparing assemblies, but it should be interpreted alongside other metrics such as the number of contigs, the total assembly length, and the completeness and contamination estimates.

How long does it take to annotate a bacterial genome?

Prokka can fully annotate a draft bacterial genome in about 10 minutes on a typical desktop computer [<a href="#ref-1">1</a>]. The annotation time depends on the genome size, the number of contigs, and the computational resources available. Prokka produces standards compliant output files for further analysis or viewing in genome browsers [<a href="#ref-1">1</a>].

What should I do if my assembly fails?

First, check the raw read quality and the trimming parameters. Poor quality input data is the most common cause of assembly failure. Second, check for contamination and remove contaminated reads if present. Third, adjust the assembly parameters, particularly the k-mer size. Fourth, consider using a different assembler or a different assembly mode. If the assembly continues to fail, escalate to a bioinformatics specialist with the assembly records and raw data files.

Can I assemble a bacterial genome on a desktop computer?

Yes, bacterial genome assembly can run on a desktop computer. SPAdes can assemble a typical bacterial genome in minutes to hours on a modern desktop computer. Long-read assembly with PacBio data has been successfully performed on a slightly improved desktop computer, closing six Dickeya solani genomes [<a href="#ref-2">2</a>]. The computational requirements depend on the genome size, coverage, and read length, but most bacterial assemblies are feasible on standard hardware.

Related Bacteriology Guides

• Avian Cholera (Fowl Cholera): Etiology, Pathogenesis, and Control
• Bacterial Pathogens in Poultry Meat: Etiology, Risks, and Control
• Fowl Cholera in Poultry: Etiology, Diagnosis, and Control
• Poultry Diseases Caused by Bacteria: Pathogenesis and Management Strategies
• Bacterial Pathogens in Poultry Meat: Salmonella, E. coli, Campylobacter, and Control Strategies

References and Further Reading

References

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