Metagenomics Pipeline: From Raw Reads to Taxonomic and Functional Profiles

Metagenomics analysis transforms raw sequencing data into taxonomic and functional profiles of microbial communities. This article describes a modular pipeline covering quality control, host removal, assembly, binning, taxonomic classification, and functional annotation, with tool recommendations and practical decision criteria for each stage. The intended readers are students, researchers, analysts, and life-science professionals who need a structured approach to metagenomic data analysis and want to understand the choices, limitations, and quality controls at each step.

At a Glance

The table below summarizes the main pipeline stages, representative tools, primary outputs, and key decision points. Tool selection depends on your research question, computational resources, and data type.

Scope and Context of Metagenomic Analysis

Shotgun metagenomic sequencing captures DNA from all organisms present in a sample, enabling simultaneous investigation of taxonomic composition, functional potential, genome-resolved diversity, and strain-level variation. This approach differs from amplicon sequencing, which targets specific marker genes such as the 16S rRNA gene for bacteria and archaea or ITS for fungi. Amplicon pipelines like EasyAmplicon provide integrated modules for quality control, merging of paired-end reads, dereplication, clustering or denoising, chimera detection, feature table generation, taxonomic diversity analysis, compositional analysis, biomarker discovery, and publication-quality visualization. EasyAmplicon is cross-platform, open-source, and community-supported, with more than 30 modules and R packages integrated into RStudio, and the entire analysis for dozens of samples can be completed within three hours on an ordinary laptop.

Shotgun metagenomics provides broader coverage than amplicon sequencing because it samples all genes in all organisms within a sample. However, the analysis is computationally more demanding and requires careful attention to data quality, reference database selection, and interpretation of results. The complexity of metagenomic study warrants following the Findability, Accessibility, Interoperability, and Reusability (FAIR) Guiding Principles for scientific data management to ensure that scientific findings can be more easily validated by the research community.

Core Principles of Pipeline Design

A well-designed metagenomics pipeline is modular, reproducible, and transparent. Modularity allows you to replace individual components without rebuilding the entire workflow. Reproducibility ensures that the same inputs produce the same outputs regardless of when or where the analysis runs. Transparency means that each step, parameter, and database version is documented and accessible.

Workflow managers such as Nextflow and Snakemake support these principles through containerized execution and automated provenance tracking. For example, zAMP is implemented as a Snakemake workflow that runs within Singularity or Docker containers and is installable via Bioconda. Shotgun-NF is a modular Nextflow DSL2 workflow that integrates read-based taxonomic and functional profiling, metagenome assembly, genome binning, metagenome-assembled genome quality assessment, abundance estimation, genome annotation, antimicrobial resistance detection, biosynthetic gene cluster prediction, and strain-level comparative analysis within a unified framework. It uses containerized execution with Conda and Apptainer, facilitating portability across local workstations, high-performance computing environments, and cloud infrastructures. Reproducibility is supported through automated provenance tracking, including execution reports, runtime traces, workflow timelines, database manifests, and directed acyclic graph visualizations.

Educational pipelines can help you understand the underlying logic. TaxoFlow is an open, interactive, and web-based tutorial that guides scholars with basic command-line skills through the detailed development of a validated and reproducible Nextflow metagenomics classification pipeline. It integrates Bowtie2 for host sequence removal, Kraken2 for taxonomic classification, and Bracken for species abundance re-estimation, with an emphasis on simplicity, modularity, and containerization.

Data Inputs and Experimental Design Considerations

Sequencing Platform and Read Length

The choice of sequencing platform affects downstream analysis decisions. Illumina paired-end reads are the most common input for shotgun metagenomics pipelines. Read length and depth influence assembly quality, taxonomic resolution, and functional annotation completeness. Longer reads generally improve assembly contiguity but may have higher error rates depending on the platform.

Sample Type and Community Complexity

Soil metagenomes present particular challenges due to high complexity and diversity. The National Ecological Observatory Network (NEON) performs annual shotgun sequencing of soils at 47 sites across the United States and provides open data and programming tutorials. Processing raw soil metagenome sequencing reads requires a bioinformatics pipeline tailored to the high complexity and diversity of the soil microbiome, including cleaning raw reads, taxonomic classification, assembly into contigs or genomes, annotation of predicted genes using custom protein databases, and exporting data for downstream analysis.

Biofilm metagenomics adds another layer of complexity because biofilms consist of diverse microbes including bacteria, viruses, and fungi that work together to enhance colonization of surfaces and produce essential nutrients. Metagenomic sequencing enables comprehensive sampling of all genes in all organisms present within a biofilm sample, but the complexity of biofilm metagenomic study warrants careful attention to data management and workflow reproducibility.

Mock Communities and Controls

Mock community samples with known composition are valuable for assessing pipeline performance. A 2024 study in Scientific Data assessed publicly available shotgun metagenomics processing packages including bioBakery, Just a Microbiology System (JAMS), Whole metaGenome Sequence Assembly V2 (WGSA2), and Woltka using 19 publicly available mock community samples and a set of five constructed pathogenic gut microbiome samples. The study used the Aitchison distance, a sensitivity metric, and total False Positive Relative Abundance for accuracy assessments. Overall, bioBakery4 performed best with most accuracy metrics, while JAMS and WGSA2 had the highest sensitivities. This work provides an unbiased assessment of shotgun metagenomics packages and presents results assessing performance using mock community sequence data.

Quality Control and Preprocessing

Raw Read Assessment

Quality control begins with examining raw sequencing reads for base quality scores, adapter contamination, GC content, and duplication levels. FastQC provides per-sample quality reports, and MultiQC aggregates these reports across many samples for comparison. These tools help you identify problematic samples before downstream analysis.

Trimming and Filtering

Trimming removes low-quality bases and adapter sequences. The specific trimming parameters depend on your sequencing platform and the quality distribution of your reads. Overly aggressive trimming can remove useful sequence data, while insufficient trimming can introduce errors into downstream analyses. After trimming, you should reassess read quality to confirm that the filtering achieved the intended effect.

Records and Measurements

Maintain a quality control log for each sample that records the number of raw reads, the number of reads after trimming, the percentage of reads retained, and the quality metrics before and after processing. This log supports reproducibility and helps you identify samples that fail quality thresholds. If a sample has unusually low read retention or poor quality scores, investigate whether the issue stems from the sequencing run, the sample preparation, or the trimming parameters.

Host Removal and Read Filtering

Rationale for Host Removal

Host DNA can constitute a large fraction of sequencing reads in clinical or environmental samples. Removing host sequences reduces the computational burden of downstream analysis and prevents host reads from being misclassified as microbial. The viGEN pipeline for viral RNA detection in human tumors illustrates this approach: the first module aligns and filters out human RNA sequences, and the second module maps the remaining unaligned reads against reference genomes of all known and sequenced human viruses.

Tools and Approaches

Bowtie2 is commonly used for aligning reads to a host reference genome and retaining unmapped reads for downstream analysis. Kraken2 can also be used for taxonomic classification and alignment of sequencing reads. The TaxoFlow tutorial integrates Bowtie2 for host sequence removal followed by Kraken2 for taxonomic classification.

Decision Criteria

Choose the host reference genome that matches your sample source. For human samples, use a human reference genome. For environmental samples, host removal may be unnecessary unless you are studying a specific host-associated community. Document the reference genome version and the alignment parameters used, because these choices affect which reads are retained.

Taxonomic Classification

K-mer-Based Methods

K-mer-based classifiers such as Kraken2 and CLARK are fast because they compare k-mers from sequencing reads against a database of k-mers from known genomes. Kraken2 is widely used and can be combined with Bracken to re-estimate species abundance. The LongStrain pipeline uses Kraken2 for taxonomic classification and Bowtie2 for alignment of sequencing reads in a strain-level analysis approach.

Alignment-Based Methods

Alignment-based methods such as DIAMOND align reads against protein or nucleotide databases. These methods are generally slower than k-mer-based approaches but can provide higher sensitivity for divergent sequences. The UGENE workflow designer allows graphical wrapping of Kraken, CLARK, and DIAMOND, combinable serially and in parallel through the workflow designer, with multiple customizable reference databases. Ensemble classification voting is available through the WEVOTE algorithm, with augmented output in the form of detailed table reports.

Phylogenetic Methods

Phylogenetic methods use full-length read alignments and phylogenetic analysis to classify samples. MGtree combines full-length read alignments with phylogenetic analysis and accurately genotypes viral samples from challenging norovirus and HPV datasets. MGtree outperforms Kraken2 and Centrifuge and succeeds with low-input samples where de novo assembly fails. Its correct assignments across highly mutant and coinfected samples highlight its ability to resolve viral genotypes and improve classification precision in complex samples.

Database Selection

The choice of reference database substantially affects classification results. Databases differ in their taxonomic scope, sequence diversity, and update frequency. A database that lacks representation of certain taxa will fail to classify reads from those taxa. Document the database version and the date of download to support reproducibility.

Performance Assessment

Mock community data provide a basis for comparing classification performance across pipelines. The 2024 Scientific Data study used sensitivity metrics and false positive relative abundance to assess accuracy. When you evaluate a classifier, consider both sensitivity and precision, because a classifier that reports many false positives can mislead downstream interpretation.

Assembly

Why Assemble?

Assembly reconstructs longer contiguous sequences from short reads. Assembled contigs provide genomic context for functional annotation and enable genome binning. However, assembly is computationally intensive and may fail for highly complex communities or low-abundance organisms.

Assembler Selection

MetaSPAdes, MEGAHIT, and IDBA-UD are commonly used metagenomic assemblers. Each has different memory requirements and performance characteristics. The choice of assembler depends on your community complexity, sequencing depth, and available computational resources. The NEON soil metagenome workflow describes assembly into contigs or genomes as part of a pipeline tailored to the high complexity and diversity of the soil microbiome.

Assembly Quality Assessment

After assembly, assess contig length distributions, N50 statistics, and the number of contigs. These metrics indicate assembly contiguity but do not directly measure correctness. Check for misassemblies by examining read coverage and alignment consistency. If assembly quality is poor, consider adjusting parameters or using a different assembler.

Binning and Metagenome-Assembled Genomes

Binning Approaches

Binning groups contigs into bins that represent putative microbial genomes. Tools such as MetaBAT2, MaxBin2, and CONCOCT use sequence composition and coverage information to cluster contigs. The resulting metagenome-assembled genomes (MAGs) enable genome-resolved analysis of community members.

Quality Assessment of MAGs

Assess MAG quality using completeness and contamination estimates. Completeness indicates the fraction of expected single-copy genes present in the bin, and contamination indicates the presence of sequences from multiple organisms. The IMG/M data management and analysis system contains annotated isolate genome and metagenome datasets and includes a tool for statistical analysis of mixed sets of genomes and metagenome bins. IMG/M is available at https://img.jgi.doe.gov/m/ and contains datasets sequenced at the DOE Joint Genome Institute, submitted by external users, or imported from public sources such as NCBI.

Decision Criteria for MAG Use

Define thresholds for completeness and contamination based on your research question. High-quality MAGs suitable for detailed genomic analysis require high completeness and low contamination. Lower-quality bins may still be useful for abundance estimation or presence-absence analysis. Document the thresholds you apply and the number of MAGs that pass each threshold.

Functional Annotation

Gene Prediction and Annotation

Functional annotation predicts genes in assembled contigs or MAGs and assigns functional categories based on reference databases. Prokka provides rapid prokaryotic genome annotation, and eggNOG-mapper assigns functional categories using the eggNOG database. The IMG/M system processes new datasets with the prokaryotic annotation pipeline v.5, which includes extended protein family assignments.

Functional Profiling from Reads

Read-based functional profiling assigns functional categories directly to sequencing reads without assembly. This approach avoids the computational cost of assembly and can detect functions from organisms that are too low in abundance to assemble. However, read-based profiling may be less sensitive for functions that require full-length gene context.

Database Considerations

Functional annotation results depend heavily on the reference database used. Different databases have different coverage of protein families, metabolic pathways, and taxonomic groups. The NEON soil metagenome workflow describes annotation of predicted genes using custom protein databases, which allows researchers to tailor the annotation to their specific research questions.

Connecting Function to Taxonomy

A key limitation of many functional annotation approaches is that they separate function from taxonomy. The META-DIFF pipeline addresses this limitation by detecting differentially abundant k-mers independently of any prior annotation, then reconstructing longer sequences from those k-mers and providing biological context. This approach can identify differentially abundant elements, including unknown DNA, prior to annotation, and provides users with exploratory power to unravel microbiome changes.

Strain-Level Analysis

Rationale for Strain-Level Resolution

Studying within-species variations enhances the understanding of microbial biological processes. Most existing methods designed for strain-level analysis lack the capability to concurrently assess both strain proportions and genome-wide single nucleotide variants (SNVs) across longitudinal metagenomic samples. LongStrain is an integrated pipeline for the analysis of large-scale metagenomic data from individuals with longitudinal or repeated samples. It uses Kraken2 and Bowtie2 for taxonomic classification and alignment of sequencing reads, then jointly models strain proportions and shared haplotypes across samples within individuals. This approach specifically targets tracking a primary strain and a secondary strain for each subject, providing their respective proportions and SNVs as output.

Applications and Limitations

Strain-level analysis is valuable for tracking transmission, monitoring pathogen evolution, and understanding microbial dynamics over time. However, strain-level resolution requires sufficient sequencing depth and may not be achievable for all taxa in a community. The LongStrain study demonstrated that the pipeline is superior to two genotyping methods and two deconvolution methods across a majority of simulated scenarios, and illustrated potential applications in real data analysis of The Environmental Determinants of Diabetes in the Young study and a gastric intestinal metaplasia microbiome study.

Differential Abundance Analysis

Traditional Approaches

Traditional case-control metagenomic studies depend on taxonomic and functional databases. Because annotation occurs before differential analysis, these approaches are limited to known elements and keep function and taxonomy separate. Binning strategies can reconstruct genomes to mitigate this issue, but they require an assembly step, preventing the use of all available sequencing data.

K-Mer-Based Differential Analysis

META-DIFF is a pipeline based on differentially abundant k-mers independently of any prior annotation. From those k-mers, it reconstructs longer sequences and provides biological context, as well as the best set of unitigs to discriminate between conditions. Across both taxonomy-centric and functionally-centric benchmarks, it showed robust performance and displayed great reproducibility. It behaved more conservatively than other univariate methodologies, maintaining high precision at the expense of recall, particularly in conditions of low fold-change and limited sequencing depth. The efficacy of META-DIFF was further validated through its application to a real-world colorectal cancer dataset, which produced both confirmatory and novel results compared with those of previous publications.

Workflow Implementation Options

Command-Line Pipelines

Command-line pipelines provide maximum flexibility and control over each analysis step. Tools such as Kraken2, Bowtie2, and MEGAHIT are command-line programs that can be combined in shell scripts or workflow managers. This approach requires familiarity with the command line and careful documentation of parameters.

Workflow Managers

Workflow managers such as Nextflow and Snakemake provide structure, reproducibility, and scalability. They handle dependency management, parallel execution, and provenance tracking. Shotgun-NF is a modular Nextflow DSL2 workflow for end-to-end shotgun metagenomics analysis from paired-end Illumina sequencing data. zAMP is a Snakemake workflow that runs within Singularity or Docker containers and is installable via Bioconda. The TaxoFlow tutorial emphasizes simplicity, modularity, and containerization, empowering users with both conceptual understanding and practical implementation skills.

Graphical Interfaces

Graphical interfaces lower the barrier to entry for users who are not comfortable with the command line. UGENE is a free, open-source, cross-platform bioinformatics software that deploys pre-defined pipelines and a flexible instrument to design new workflows and visually build multi-step analytics pipelines. The new UGENE v.1.31 release offers graphical, user-friendly wrapping of a number of popular command-line metagenomics classification programs, combinable serially and in parallel through the workflow designer, with multiple customizable reference databases.

Cloud-Based Platforms

Cloud-based platforms provide access to scalable computing resources without direct management overhead. The Genomic Data Science Analysis, Visualization, and Informatics Lab-space (AnVIL) provides a cloud-based platform for democratizing access to large-scale genomics data and analysis tools. The AnVILWorkflow R package enables the convenient execution of bioinformatics workflows hosted on AnVIL directly from an R environment, simplifying the setup of the cloud computing environment, input data formatting, workflow submission, and retrieval of results through intuitive functions. AnVILWorkflow has been demonstrated for metagenomics analysis with bioBakery, among other use cases.

Dockerized Workflows

Dockerized workflows package software and dependencies into containers, ensuring that the analysis runs identically across different computing environments. A dockerized, self-learning bioinformatics workflow was proposed to increase the community adoption of metagenomics toolkits in metagenomics and meta-transcriptomics investigation. Containerization addresses the challenges of software dependencies and version conflicts that complicate metagenomic analysis.

Records and Measurements

Documentation Requirements

Maintain detailed records of every analysis step, including software versions, database versions, parameter settings, and input file paths. This documentation supports reproducibility and allows you to trace the origin of any result. The Shotgun-NF workflow supports this through automated provenance tracking, including execution reports, runtime traces, workflow timelines, database manifests, and directed acyclic graph visualizations.

Quality Metrics to Record

Record quality metrics at each pipeline stage, including read counts, read retention rates, assembly statistics, bin completeness and contamination estimates, and classification rates. These metrics help you identify problems and compare results across samples or studies.

Metadata Standards

Standardized metadata is essential for interpreting metagenomic results and for data sharing. The FAIR Guiding Principles emphasize that data should be findable, accessible, interoperable, and reusable. A 2026 article in Biology and Fertility of Soils emphasizes the need for standardized metadata, data availability, and workflow reproducibility in metagenomic studies, and encourages authors to connect microbial functional potential to genomic context, functional redundancy, and taxonomic diversity.

Common Failure Patterns

Inadequate Quality Control

Skipping or rushing quality control can introduce errors that propagate through the entire pipeline. Low-quality reads can cause misclassification, assembly errors, and incorrect functional annotations. Always assess read quality before and after trimming.

Database Mismatches

Using a reference database that does not match your sample type can lead to poor classification rates or false assignments. For example, a database lacking representation of soil microorganisms will fail to classify many soil metagenome reads. Choose databases that match your expected community composition and document the database version.

Overinterpretation of Results

Metagenomic results are estimates, not absolute measurements. Taxonomic classification depends on database representation, and functional annotation depends on reference database coverage. The 2026 Biology and Fertility of Soils article highlights opportunities and limitations of current tools and workflows utilized for the analyses of metagenomic sequencing data, providing guidance for methodological thoroughness and transparency.

Computational Resource Underestimation

Metagenomic assembly and classification can require substantial memory and processing time. Underestimating resource requirements can cause pipeline failures or excessively long runtimes. Test your pipeline on a small subset of data before running the full dataset.

Ignoring Reproducibility

Failure to document parameters, versions, and environments makes it impossible to reproduce or validate results. Use workflow managers with containerization and provenance tracking to support reproducibility.

Limitations and Interpretation Constraints

Database Dependency

All taxonomic and functional classification methods depend on reference databases. Sequences from organisms not represented in the database will remain unclassified or be assigned to the nearest known relative. This limitation is inherent to database-dependent approaches and should be acknowledged in interpretations.

Resolution Limits

Taxonomic resolution varies across lineages and depends on the marker genes or genomic regions used. Some closely related species cannot be distinguished by short reads. Strain-level analysis requires specialized approaches such as those implemented in LongStrain.

Functional Potential versus Activity

Metagenomics reveals the genetic potential of a community, not its actual activity. The presence of a gene does not confirm that it is expressed or that the corresponding function is active. Metatranscriptomics can complement metagenomics by measuring gene expression.

Assembly Limitations

Assembly may fail to reconstruct genomes from low-abundance organisms or highly complex communities. The MGtree study notes that de novo assembly fails for some low-input samples, and the pipeline succeeds where assembly fails by using full-length read alignments and phylogenetic analysis.

Safety and Regulatory Context

Data Sharing Policies

Genomic data may be subject to data sharing policies that govern how sequence data can be stored, shared, and used. The NIH Genomic Data Sharing Policy at https://sharing.nih.gov/genomic-data-sharing-policy provides requirements for data sharing in NIH-funded research. Researchers should be aware of applicable policies before depositing or sharing metagenomic data.

Data Repositories

Public data repositories such as NCBI at https://www.ncbi.nlm.nih.gov/ provide access to reference genomes, sequence data, and analysis tools. The IMG/M system at https://img.jgi.doe.gov/m/ contains annotated isolate genome and metagenome datasets sequenced at the DOE Joint Genome Institute, submitted by external users, or imported from public sources such as NCBI.

Training and Education

Training resources are available from official sources. The EMBL-EBI Training portal at https://www.ebi.ac.uk/training provides courses and materials for bioinformatics and metagenomics analysis. The NEON soil metagenome dataset serves as a valuable educational resource, thanks to its open data and programming tutorials.

Professional Escalation Criteria

When to Seek Expert Assistance

Consider consulting a bioinformatics specialist or collaborating with an experienced metagenomics analyst when you encounter any of the following situations:

• Your samples have unusual community composition that may not be represented in standard reference databases
• You need strain-level resolution for clinical or epidemiological decision-making
• Your assembly or binning results are poor despite following recommended parameters
• You are unsure whether your quality control thresholds are appropriate for your data type
• You need to comply with specific data sharing or regulatory requirements
• Your results will inform clinical, public health, or regulatory decisions

Validation Before Escalation

Before escalating, verify that you have documented all parameters and versions, checked quality metrics at each stage, and compared your results with mock community or control samples where available. This information helps the expert diagnose the issue efficiently.

Frequently Asked Questions

What is the difference between amplicon and shotgun metagenomics pipelines?

Amplicon pipelines target specific marker genes such as 16S rRNA for bacteria and archaea or ITS for fungi. EasyAmplicon is an example of an amplicon pipeline that includes data quality control, merging of paired-end reads, dereplication, clustering or denoising, chimera detection, generation of feature tables, taxonomic diversity analysis, compositional analysis, biomarker discovery, and publication-quality visualization. Shotgun metagenomics pipelines analyze all DNA in a sample, providing broader coverage of taxonomic composition, functional potential, genome-resolved diversity, and strain-level variation. Shotgun analysis is computationally more demanding but provides more comprehensive information.

How do I choose between k-mer-based and alignment-based taxonomic classifiers?

K-mer-based classifiers such as Kraken2 and CLARK are fast and memory-efficient, making them suitable for large datasets. Alignment-based methods such as DIAMOND are slower but can provide higher sensitivity for divergent sequences. The choice depends on your sample size, computational resources, and the expected diversity of your community. Some pipelines combine both approaches, and the UGENE workflow designer allows serial and parallel combination of Kraken, CLARK, and DIAMOND with multiple customizable reference databases.

What is the role of mock communities in validating a metagenomics pipeline?

Mock communities with known composition provide a ground truth for assessing pipeline accuracy. A 2024 study in Scientific Data used 19 publicly available mock community samples and five constructed pathogenic gut microbiome samples to assess bioBakery, JAMS, WGSA2, and Woltka, using the Aitchison distance, a sensitivity metric, and total False Positive Relative Abundance for accuracy assessments. Running your pipeline on mock community data helps you understand its strengths and limitations before applying it to real samples.

When should I use assembly and binning instead of read-based profiling?

Assembly and binning are appropriate when you need genome-resolved information, such as the genomic context of functional genes or the reconstruction of individual genomes from a community. Read-based profiling is faster and can detect functions from organisms that are too low in abundance to assemble. The NEON soil metagenome workflow describes assembly into contigs or genomes as part of a pipeline tailored to the high complexity and diversity of the soil microbiome.

How do I assess the quality of metagenome-assembled genomes?

Assess MAG quality using completeness and contamination estimates. Completeness indicates the fraction of expected single-copy genes present in the bin, and contamination indicates the presence of sequences from multiple organisms. Define thresholds based on your research question, and document the number of MAGs that pass each threshold. The IMG/M system provides tools for statistical analysis of mixed sets of genomes and metagenome bins.

What are the main limitations of functional annotation from metagenomic data?

Functional annotation depends on reference databases, and sequences from organisms not represented in the database will remain unannotated or be assigned to the nearest known relative. Metagenomics reveals genetic potential, not actual activity, so the presence of a gene does not confirm that it is expressed. The META-DIFF pipeline addresses some limitations by detecting differentially abundant k-mers independently of any prior annotation, which can identify unknown DNA prior to annotation.

How can I make my metagenomics pipeline reproducible?

Use workflow managers such as Nextflow or Snakemake with containerization to manage software dependencies and versions. Document all parameters, database versions, and input file paths. The Shotgun-NF workflow supports reproducibility through automated provenance tracking, including execution reports, runtime traces, workflow timelines, database manifests, and directed acyclic graph visualizations. The TaxoFlow tutorial emphasizes simplicity, modularity, and containerization to empower users with both conceptual understanding and practical implementation skills.

What should I do if my taxonomic classification rates are low?

Low classification rates can result from database mismatches, poor read quality, or the presence of organisms not represented in the reference database. Check your quality control metrics, verify that your database matches your sample type, and consider using a different classifier or database. The MGtree pipeline uses full-length read alignments and phylogenetic analysis to classify samples and succeeds with low-input samples where de novo assembly fails, which may be useful for challenging samples.

Related Bioinformatics Guides

• Metagenomics Taxonomic Classification: Kraken2 and Functional Annotation Pipelines
• Qiime2 Taxonomic Classification: Structural Analysis and Computational Methodologies in Bioinformatics
• Ribosomal RNA (rRNA): Structure, Function, and Taxonomic Profiling in Metagenomics
• AlphaFold2-Based Structural Modeling and Functional Annotation of PRRSV Nonstructural Proteins
• Structural and Functional Annotation of Novel Bat Coronaviruses using AlphaFold2 and Molecular Docking

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.
• An integrated strain-level analytic pipeline utilizing longitudinal metagenomic data.. Microbiology spectrum, 2024.
• Biofilm marker discovery with cloud-based dockerized metagenomics analysis of microbial communities.. Briefings in bioinformatics, 2024.
• Flexible design of multiple metagenomics classification pipelines with UGENE.. Bioinformatics (Oxford, England), 2019.
• EasyAmplicon: An easy-to-use, open-source, reproducible, and community-based pipeline for amplicon data analysis in microbiome research.. iMeta, 2023.
• The National Ecological Observatory Network's soil metagenomes: assembly and basic analysis.. F1000Research, 2021.
• viGEN: An Open Source Pipeline for the Detection and Quantification of Viral RNA in Human Tumors.. Frontiers in microbiology, 2018.
• The IMG/M data management and analysis system v.6.0: new tools and advanced capabilities.. Nucleic acids research, 2021.
• zAMP and zAMPExplorer: reproducible scalable amplicon-based metagenomics analysis and visualization.. Bioinformatics advances, 2025.
• Shotgun-NF: A reproducible Nextflow pipeline for end-to-end shotgun metagenomics analysis. 2026.
• MGtree: A Fast and Flexible Alignment-Based Metagenomics Pipeline.. 2026.
• TaxoFlow: A Step-by-Step Tutorial to Build a Nextflow Pipeline for Metagenomics Taxonomic Classification. 2026.
• META-DIFF: a k-mer-based pipeline that detects differentially abundant sequences in metagenomics whole genome sequencing.. 2026.
• TaxoFlow: The Tutorial. An Educational Nextflow Pipeline for Metagenomics Taxonomic Profiling. 2025.
• Mock community taxonomic classification performance of publicly available shotgun metagenomics pipelines. Scientific Data, 2024.
• Digging deeper into soil metagenomics: Opportunities and limitations for studying the genomic potential of soil bacteria and archaea. Biology and Fertility of Soils, 2026.
• An Integrated Genomics Workflow Tool: Simulating Reads, Evaluating Read Alignments, and Optimizing Variant Calling Algorithms. International Conference on Bioinformatics & Computational Biology, 2024.
• AnVILWorkflow: A runnable workflow package for Cloud-implemented bioinformatics analysis pipelines. Research Square, 2024.
• AnVILWorkflow: A runnable workflow package for Cloud-implemented bioinformatics analysis pipelines. F1000Research, 2024.
• Personalized Constraint-Based Modeling of Microbial Communities from Metagenomic Data. Methods in Molecular Biology, 2026.

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