Shotgun Metagenomics · Untargeted Community Profiling

Sequence the whole community. Taxonomy and function.

End-to-end Shotgun Metagenomics sequencing and analysis for microbial community characterisation — taxonomic profiling across bacteria, archaea, fungi and viruses alongside functional characterisation of the community.

Plan Your Project → See the Workflow
98.7%
Mean host-read removal (demo)
~185
Mean species per sample (demo)
~92%
Mean non-host reads retained
MAG
Optional genome recovery
About the Service

Untargeted community profiling beyond amplicons

Nucleome Informatics offers end-to-end Shotgun Metagenomics (SGM) sequencing and data analysis services for microbial community characterisation across environmental, clinical, agricultural and industrial samples.

Unlike amplicon-based approaches, shotgun metagenomics sequences all DNA present in a sample, enabling taxonomic profiling across bacteria, archaea, fungi and viruses alongside functional characterisation of the community.

Our analysis combines rigorous sequencing quality control, host and contaminant read removal, taxonomic classification, functional profiling and, where required, metagenome assembly. Analyses use established, widely benchmarked tools and curated reference databases suited to the sample type and study objectives.

  • Sequencing data quality assessment and preprocessing
  • Host DNA depletion and contaminant read removal
  • Taxonomic classification at multiple taxonomic ranks
  • Functional profiling of genes and metabolic pathways
  • Optional metagenome assembly and genome (MAG) recovery
  • Alpha- and beta-diversity analysis
  • Comparative and differential abundance analysis across groups
  • Secure delivery of raw, intermediate and final analysis files

Beyond Taxonomy

Shotgun sequencing captures the full genetic content of the community — enabling gene family and pathway-level functional profiling alongside species composition.


Domains Covered

Bacteria Archaea Fungi Viruses
Learning Path

From first principles to expert-level analysis

New to metagenomics? Start with Foundations. Planning a study or reviewing results? Jump to Practitioner or Expert. Each level builds on the one before it, so you can read this page as a short course or dip in where you need to.

The big idea in plain language: a microbial community is a mixture of many organisms. Shotgun metagenomics reads random fragments of all the DNA in the mixture, so you learn both who is there and what genes they carry.

LEVEL 1

A metagenome is all the DNA in a sample

Stool, soil or water holds thousands of microbes at once. The metagenome is the combined genetic material of all of them, bacteria, archaea, fungi and viruses together.

LEVEL 1

Shotgun means random fragments

The DNA is broken into short pieces and every piece is sequenced. Unlike amplicon methods, no single marker gene is chosen in advance, so nothing is missed by primer choice.

LEVEL 1

Reads are matched to reference genomes

Each read is compared with databases of known microbial genomes. The best match tells which organism, or which gene, the read probably came from.

LEVEL 1

Host DNA is removed first

Samples from people or animals contain host DNA that can swamp the microbial signal. Reads that match the host genome are filtered out so the sequencing effort is spent on microbes.

LEVEL 1

You get two kinds of answers

Taxonomy says which species are present and how abundant they are. Function says which genes and metabolic pathways the community can carry out.

LEVEL 1

Diversity summarises the whole community

Alpha diversity describes one sample (how rich and even it is). Beta diversity compares samples (how different their communities are).

For scientists planning a project or reading a report: how each step works and how it can mislead.

LEVEL 2

Quality control comes first

Reads are checked for base quality (Q30 means a 1-in-1,000 error chance), adapters are trimmed and poor reads are dropped, so errors do not turn into false taxa.

LEVEL 2

Classifiers work in different ways

k-mer tools such as Kraken2 look up short words in a database, marker-gene tools such as MetaPhlAn count clade-specific genes, and alignment tools map reads to genomes. Each has its own speed, sensitivity and bias.

LEVEL 2

The database sets the limits

A microbe missing from the reference cannot be named. Databases also differ in size and curation, so results depend on which one suits the sample type.

LEVEL 2

Function from gene families and pathways

Reads are assigned to gene families, and gene families are rolled up into pathways. Tools such as HUMAnN also report which species contribute to each pathway.

LEVEL 2

Assembly builds longer pieces

Overlapping reads can be stitched into contigs, and contigs with similar composition and coverage are grouped (binned) into metagenome-assembled genomes (MAGs) for organisms never cultured.

LEVEL 2

Depth decides what you can see

Rare species need more reads to be detected. Highly diverse samples such as soil need far more depth than a simple community, so the sequencing target should match the sample type.

For method developers and reviewers: the statistical and technical caveats that decide whether an apparent result holds up.

LEVEL 3

Abundances are compositional

Sequencing gives proportions, not absolute amounts. If one taxon blooms, all others fall in relative terms even though they did not change. Log-ratio methods address this (Gloor et al., 2017).

LEVEL 3

Differential abundance methods disagree

A comparison on 38 datasets found that different methods identify different sets of significant taxa, so conclusions that hold across several methods are more trustworthy (Nearing et al., 2022).

LEVEL 3

Reads can be misassigned

Reads from organisms missing from the database are often assigned to their nearest relative. Confidence thresholds and abundance re-estimation, as in Bracken, reduce but do not remove this (Wood et al., 2019; Lu et al., 2017).

LEVEL 3

Low-biomass samples need controls

Reagent and environmental contaminants dominate when little microbial DNA is present. Extraction blanks and negative controls are required to tell contamination from signal (Salter et al., 2014; Eisenhofer et al., 2019).

LEVEL 3

MAG quality must be measured

Bins can mix genomes or miss parts. Standards classify MAGs by completeness and contamination, estimated with tools such as CheckM2; high quality means at least 90% complete and under 5% contaminated (Bowers et al., 2017).

LEVEL 3

Benchmarks guide tool choice

The CAMI challenges evaluate assemblers, binners and profilers on shared data with known answers, showing that performance varies with community complexity and strain diversity (Sczyrba et al., 2017; Meyer et al., 2022).

Metagenome
The combined genetic material of all organisms in a sample.
Read
One sequenced DNA fragment, typically 100–150 bases from short-read platforms.
Host depletion
Removing reads that match the host genome before microbial analysis.
k-mer
A short word of k bases, used as a lookup key to classify reads.
LCA
Lowest common ancestor: the taxonomic rank shared by all matches of a read.
Contig
A longer sequence assembled from overlapping reads.
Binning
Grouping contigs that appear to come from the same organism.
MAG
Metagenome-assembled genome: a draft genome recovered from a bin.
Alpha diversity
Richness and evenness inside one sample, e.g. the Shannon index.
Beta diversity
Difference in community composition between two samples.
Compositional data
Data where only proportions are known, so values depend on each other.
Pathway
A series of enzyme-catalysed steps, such as the synthesis of a vitamin.
The Science

How a mixed community becomes reads and profiles

What shotgun sequencing captures that amplicons miss, how reads are assigned to organisms, and why sequencing depth matters.

Figure 1 Amplicon versus shotgun sequencing schematic
Amplicon: one marker gene per organism Primers amplify only the marker gene (orange). Bacteria and archaea only; limited to genus-level resolution; no direct gene content. Shotgun: random fragments of everything Fragments sampled across every genome (green). All domains including fungi and viruses; genes and pathways; genome recovery. What you can answer Amplicon: Who is there (mostly bacteria and archaea), and how abundant? Shotgun: Who is there, what can they do, and which genomes can be recovered? Cost per sample is lower and depth needs are smaller. Cost is higher and more depth is needed, especially with host DNA.
Figure 1 | What each method reads. Amplicon sequencing targets a marker gene such as 16S rRNA. Shotgun sequencing reads random fragments across all genomes, including viruses and fungi, which lets taxonomy be placed alongside gene content (Quince et al., 2017).
Figure 2 How k-mer classification finds a read's taxon schematic
1 · Cut the read into k-mers ACGTTGCAGGTCAAGTCCATTAGCGGATC… k-mer 1 k-mer 2 k-mer 3 k-mer 4 k-mer 5 2 · Look each k-mer up in the tree Order Gen. A Gen. B Sp. 1 Sp. 2 Sp. 3 3 hits The read goes to the path with most hits (Sp. 1). Mixed hits are resolved up to the shared ancestor (LCA). Database limitNovel organismsmap to a relative
Figure 2 | The k-mer idea. Each short word of the read votes for the taxa containing it. The read is assigned to the best-supported root-to-leaf path, or to a higher rank when votes conflict (Wood and Salzberg, 2014). Speed comes from using exact lookups instead of alignment.
Figure 3 Sequencing depth and detected species simulated
Figure 3 | How deep is deep enough? Species detected (at least 10 reads) against reads sequenced for three simulated communities. Simple communities level off early; highly diverse ones, like soil, keep gaining species with depth. Curves are expectations from a log-normal abundance model.
Table 1 Choosing a sequencing approach
ApproachReadsLimit
Amplicon (16S / ITS)One marker geneNo function; primer bias
Short-read shotgunAll DNA, 100–150 basesHard to resolve repeats and strains
Long-read shotgunAll DNA, kilobases longHigher cost and DNA input
MetatranscriptomicsExpressed RNAReflects activity; RNA is unstable
This service. The page describes short-read shotgun metagenomics of DNA, with optional assembly and binning. Other designs can be discussed with our team.
Table 2 Ways reads get a taxonomic label
StrategyHowTypical trait
k-mer lookupExact short words against a databaseVery fast; database-dependent
Marker genesClade-specific genesFew false positives; fewer reads used
AlignmentReads mapped to genomesSensitive; slower
Assembly firstContigs classified or binnedGenome-level; needs depth
No single best choice. Benchmarks show profilers trade sensitivity against precision, which is why the classifier and database are chosen to suit the sample type.
Technology Highlights

Built for genome-resolved communities

From host depletion through MAG recovery — a complete shotgun metagenomics toolkit.

Host & Contaminant Depletion

Removal of host and non-target reads prior to microbial analysis for accurate community profiling.

Taxonomic Profiling

Species- and genus-level classification using curated reference databases and established classifiers.

Functional Profiling

Gene family and metabolic pathway abundance estimation to characterise community function.

Assembly & MAG Recovery

Optional metagenome assembly and binning for recovery of metagenome-assembled genomes and novel organism discovery.

Diversity & Comparative Analysis

Alpha- and beta-diversity metrics with group-wise statistical comparisons.

Reproducible Reporting

Structured abundance tables, QC summaries and project reports delivered in a consistent format.

Analysis Workflow

From raw reads to community profiles

Each sample passes through a quality-controlled analysis workflow designed to move from raw sequencing data to taxonomic and functional community profiles.

Figure 4 Quality-gated metagenomics workflow infographic
STAGE 01

Data QC & pre-processing

Sequencing quality is assessed; quality filtering and preprocessing are applied.

FASTQclean FASTQ
Q30 · read counts
STAGE 02

Host & contaminant removal

Reads matching host or non-target genomes are identified and removed.

non-host reads
% host removed
STAGE 03

Taxonomic classification

Remaining reads are classified against curated microbial reference databases.

abundance tables
classified fraction
STAGE 04

Functional profiling

Genes and pathways are identified and quantified to describe functional potential.

gene familiespathways
mapped fraction
STAGE 05

Assembly & binning (optional)

Contigs are assembled and binned into metagenome-assembled genomes.

contigsMAGs
completeness · contamination
STAGE 06

Diversity & comparison

Alpha- and beta-diversity metrics; composition compared across groups.

diversitygroup tests
design reviewed
STAGE 07

Reporting & delivery

Profiles, diversity statistics and QC summaries delivered through secure storage.

reportsecure storage
access control
Figure 4 | Decision points. Amber gates mark where metrics are reviewed before the next stage. A sample with unusually low non-host read yield is flagged rather than silently carried forward.
01

Data QC & Pre-processing

Sequencing quality assessed; quality filtering and preprocessing applied.

02

Host & Contaminant Removal

Reads matching host or non-target genomes identified and removed.

03

Taxonomic Classification

Remaining reads classified against curated microbial reference databases.

04

Functional Profiling

Genes and pathways identified and quantified to characterise functional potential.

05

Assembly & Binning (Optional) — Contigs assembled and binned into metagenome-assembled genomes.

06

Diversity & Comparative Analysis — Alpha- and beta-diversity metrics; community composition compared across groups.

07

Reporting & Delivery — Taxonomic and functional profiles, diversity statistics and QC summaries delivered securely.

Assembly Options

Where study design calls for genome-resolved metagenomics, the same processed data can be assembled and binned to recover metagenome-assembled genomes, in addition to standard read-based taxonomic and functional profiling.

Quality Control & Host Depletion

Clean reads, microbes only

Good reads and a high share of microbial reads are the foundation of every later result. Host depletion matters most in samples that are mostly host DNA.

Figure 5 Read quality before and after processing simulated
Figure 5 | Per-base quality. Quality usually falls towards the end of reads (green). Trimming and filtering remove the weakest bases and reads (orange), raising the share of bases above Q30. Q30 means a 0.1% chance of an error at that base. The case study below reports mean Q30 before and after processing.
Figure 6 Why host DNA changes the sequencing plan simulated
Figure 6 | Host reads use up your budget. With 20 million reads and a gut-like community, the curve shows species detected as the host share rises. A tissue or swab sample that is 90% host leaves one tenth of the reads for microbes. Depleting host reads in the analysis recovers the microbial signal but not the lost depth.
Host depletion in practice. Reads are mapped to the host genome and those that match are removed, so abundance tables describe microbes only. Stool usually contains little host DNA, whereas swabs, saliva and tissue can contain a great deal. The share removed is reported for every sample.
Taxonomic Profiling

Who is in the community

Reads are assigned to taxa at several ranks, from phylum to species, and summarised as abundance tables.

Figure 7 Community composition across samples simulated
Figure 7 | Relative abundance. Each bar is one sample, split into two groups of 10; colours are phyla. Stacked bars give a quick overview and show sample-to-sample variability, but cannot show which differences are statistically supported.
Figure 8 The classification confidence trade-off simulated
Figure 8 | Strict versus sensitive. Reads from organisms in the database score high; reads from novel relatives get lower scores. Raising the threshold removes wrong assignments but also drops true reads. The scores are simulated to illustrate the trade-off for k-mer classifiers (Wood et al., 2019).
Figure 9 Proportions can mislead calculated example
Figure 9 | The compositional trap. Left: true amounts of three taxa. Taxon C blooms fivefold; A and B do not change. Right: relative abundance, which is all sequencing measures. A and B appear to fall even though nothing happened to them (Gloor et al., 2017).
Table 3 Reading an abundance table
QuantityMeaning
Read countReads assigned to a taxon; depends on depth
Relative abundanceShare of classified reads; sums to 100%
UnclassifiedReads with no database match
RankPhylum, genus or species; deeper ranks are less certain
Interpretation. Rare taxa near the detection limit are the least reliable. Results at species level are most dependable for well-represented genomes.
Functional Profiling

What the community can do

Reads are mapped to gene families, and gene families are rolled up into metabolic pathways, giving a view of functional potential.

Figure 10 From reads to pathways schematic
Readsnon-host Gene familiese.g. enzymes Pathwayse.g. vitamin synthesis searchcombine Who contributes? Species 1Species 2Sp. 3unclassified Stratified output splits one pathway's abundance by contributing organism. Two things are reported • Abundance: how much of the pathway's machinery was sequenced. • Coverage: how complete the pathway's steps are.
Figure 10 | Gene and pathway abundance. Reads are searched against protein databases to assign gene families; those are combined into pathways, and organism-level contributions can be traced (Beghini et al., 2021). The profile describes potential, not measured activity.
Figure 11 Pathway abundance by contributing species simulated
Figure 11 | Stratified pathways. Bars show relative abundance of eight example pathways, coloured by the species contributing. Core functions such as amino-acid synthesis are spread over many species, whereas some pathways are carried by a few. Pathway names and values are illustrative.
Assembly & Genome Recovery

From reads to draft genomes

Where study design calls for it, reads are assembled into contigs and grouped into metagenome-assembled genomes, which can reveal organisms with no reference genome.

Figure 12 Assembly and binning schematic
1 · Overlapping reads2 · Contigs3 · Bins (MAGs) MAG 1MAG 2MAG 3 Binning uses sequence composition (e.g. tetranucleotide frequency) and coverage patterns. Each MAG is checked for completeness and contamination.
Figure 12 | Genome-resolved metagenomics. Assemblers join overlapping reads into contigs; binning tools cluster contigs that share composition and abundance, ideally across several samples. Recovered MAGs expanded the known tree of life substantially (Parks et al., 2017; Pasolli et al., 2019).
Figure 13 Why binning works: composition and coverage simulated
Figure 13 | Contigs cluster by genome. Each point is a contig, plotted by GC content and log coverage. Contigs from one genome share both, so they form a cloud. Where two genomes overlap on GC, coverage separates them, which is why multi-sample coverage is valuable.
Figure 14 MAG quality assessment simulated
Figure 14 | Sorting MAGs by quality. Each point is a simulated bin. Completeness and contamination are estimated from single-copy marker genes. Shaded regions mark the standard high-quality (≥90% complete, <5% contaminated) and medium-quality (≥50%, <10%) tiers (Bowers et al., 2017).
Table 4 MAG quality tiers
TierCompletenessContaminationTypical use
High quality≥ 90%< 5%Genome-level inference, comparative genomics
Medium quality≥ 50%< 10%Taxonomy and gene content, with care
Low quality< 50%< 10%Exploratory only
Full MIMAG criteria. The formal high-quality tier also asks for rRNA and tRNA genes. The thresholds shown are the completeness and contamination parts.
Diversity Analysis

One sample, then many

Alpha diversity summarises each community; beta diversity compares communities and shows whether groups separate.

Figure 15 Alpha diversity between groups simulated
Figure 15 | Shannon diversity. Each point is a sample (n = 12 per group). The index rises with both the number of species and how evenly they are spread. The p-value is a Wilcoxon rank-sum test; with small groups a significant result needs a large difference.
Figure 16 Beta diversity ordination simulated
Figure 16 | PCoA. Each point is a sample; distance on the plot reflects community dissimilarity. Bray-Curtis (default) works on relative abundance; ticking the box switches to Aitchison distance, which respects the compositional nature of the data. The PERMANOVA test uses 999 permutations on the full distance matrix.
Comparative Analysis

Finding taxa that truly differ

Group comparisons ask which taxa differ in abundance. Because sequencing data are proportions, naive tests can flag taxa that did not change.

Figure 17 Differential abundance and compositional bias simulated
Figure 17 | Fold changes with and without centring. In this simulation 8 of 150 taxa truly increase, which lowers everyone else's relative share. Without correction, many unchanged taxa look decreased (orange); centring on the log-ratio scale re-centres them near zero. Tests are Welch t-tests on log abundance with Benjamini-Hochberg adjustment.
Table 5 Practical guidance
ConcernGood practice
Different depthsUse models that handle depth, or log-ratio methods
Many zerosFilter very rare taxa; use zero-aware approaches
Methods disagreeReport taxa found by more than one method
ConfoundersAdjust for batch, age, diet or medication where known
Absolute changeUse spike-ins or total microbial load if needed
Why this matters. The page's comparisons are group-wise statistical tests. Ask our team which method and covariates suit your design.
Deliverables

Structured outputs for metagenomics research

At project completion, clients receive a structured set of sequencing, analysis and reporting outputs via secure shared storage.

Raw & Processed Data

FASTQ files — raw and appropriately processed sequence data.

QC Reports

Sample-level sequencing and preprocessing QC summaries.

Taxonomic Profiles

Abundance tables at multiple taxonomic levels (phylum through species).

Functional Profiles

Gene family and pathway abundance tables.

Assembled Contigs / MAGs

Assembly outputs and recovered genome bins, where included in scope.

Diversity Summaries

Alpha- and beta-diversity metrics, ordination plots and group comparisons.

Summary Report

Sample-wise QC, classification and functional-profiling statistics.

Secure Delivery

Project files via secure shared storage or cloud-based access.

Sample Requirements

What we accept

Flexible input options for sequencing-plus-analysis or analysis-only projects.

DNA from Diverse Matrices

We accept genomic DNA extracted from stool, soil, water, swabs or other environmental and clinical matrices, as well as pre-generated sequencing data where analysis-only services are required.

Project-specific requirements for DNA quantity, concentration, purity, sequencing depth and read configuration are finalised according to the study design and sequencing strategy.

Contact us for a project-specific sample submission checklist and recommended sequencing depth.

Stool · Soil · Water · Swabs
or pre-generated FASTQ

Preserve DNA immediatelyFreeze samples or use a stabilisation buffer at collection. Microbial composition changes quickly at room temperature.
Use the same extraction methodDifferent kits lyse different cells. Using one method across a study keeps group differences from being extraction differences.
Add negative controlsExtraction blanks and water controls reveal reagent contamination, which matters most in low-biomass samples such as swabs and water.
Spread groups across batchesProcess cases and controls together and record extraction and sequencing batches in the sample sheet so batch effects can be checked.
Case Studies & Performance

Proven across sample types

Production metrics from gut microbiome and environmental metagenomics cohorts.

Gut Microbiome Community Profiling

Human gut · Host depletion · Taxonomic + functional

Mean raw-data Q30 91.8%
Mean processed-data Q30 96.1%
Mean host-read removal 98.7%
Mean species identified ~185
Mean non-host reads retained ~92%

Multi-sample human gut microbiome cohort processed end-to-end: sequencing QC, host read depletion, taxonomic classification and functional profiling — with high sequencing quality and effective host-read depletion.

Environmental Sample Functional Profiling

Environmental cohorts · MAG recovery

Non-host read retention Consistently high
MAG recovery Medium- to high-quality
Genome-resolved analysis Supported
Functional profiling Included
Workflow Same validated pipeline

In production environmental cohorts, we have observed consistently high non-host read retention and recovery of medium- to high-quality metagenome-assembled genomes where assembly was included in project scope.

The Frontier

Where metagenomics is heading

Metagenomics has moved from listing species to recovering genomes, tracking strains and viruses, and measuring absolute abundance. Here is what recent research and benchmarks say.

Figure 18 Layers of a shotgun dataset schematic
Reads Taxonomy · Function Genomes (MAGs) Strains · Resistance genes Viruses · Plasmids · Mobile elements Deeper layers need more depth, assembly or reference resources
Figure 18 | More from the same reads. The same data can support taxonomy and function, then strain tracking, resistance genes and viral discovery, each needing additional depth or specialised analysis.
Table 6 From standard to frontier analysis
LevelQuestion
Read-based profileWhich taxa and pathways are present?
Genome-resolvedWhich draft genomes can be recovered?
Strain-levelWhich variant of a species is present?
Resistome and viromeWhich resistance genes and viruses are carried?
Absolute abundanceHow many cells, not just what proportion?
Layered by need. The page's standard read-based profile and optional MAG recovery cover the first two rows; the rest are extensions to scope with our team.
BENCHMARKS

CAMI challenges

The second round of the Critical Assessment of Metagenome Interpretation compared assemblers, binners and profilers on realistic data with known answers, showing large differences between tools and with community complexity (Meyer et al., Nature Methods, 2022).

STATISTICS

Choosing a differential abundance method

A comparison of 14 methods on 38 datasets found that they identified different numbers and sets of significant taxa, supporting the use of several methods and consensus (Nearing et al., Nature Communications, 2022).

GENOMES

Catalogues of uncultured genomes

Reconstruction of over 150,000 genomes from 9,000 metagenomes showed that a large share of human-gut diversity lacks cultured representatives (Pasolli et al., Cell, 2019); a later catalogue of 204,938 gut genomes extended this (Almeida et al., Nature Biotechnology, 2021).

BINNING

Deep-learning binners

VAMB uses variational autoencoders to combine sequence composition and abundance across samples, recovering more high-quality genomes than earlier binners (Nissen et al., Nature Biotechnology, 2021).

QUALITY

Faster MAG quality estimation

CheckM2 uses machine learning to estimate completeness and contamination, including for novel lineages poorly covered by marker-gene sets (Chklovski et al., Nature Methods, 2023).

ABSOLUTE ABUNDANCE

Beyond proportions

Quantitative microbiome profiling adds total microbial load to relative data, and showed that load itself is a major source of variation between gut communities (Vandeputte et al., Nature, 2017).

Scope note. Nucleome's shotgun metagenomics service covers the workflow and deliverables described on this page: sequencing QC and preprocessing, host and contaminant removal, taxonomic classification at multiple ranks, gene-family and pathway profiling, optional assembly and MAG recovery, alpha- and beta-diversity analysis, and group comparisons, with raw and processed files delivered securely. The tools named in explanatory text are examples of their kind; the classifier, database and statistical method are chosen per project. Strain-level, resistome, virome, long-read, absolute-abundance and metatranscriptomic analyses are shown to place the service in context; contact our team to discuss extensions.
Why Nucleome

Service highlights

Support for research, environmental and translational metagenomics projects.

End-to-End Metagenomics

From raw sequencing data through QC, taxonomic and functional profiling, to reporting.

Robust Host Depletion

Ensures accurate community profiles free of non-target sequence.

Comprehensive Functional Insight

Gene- and pathway-level profiling alongside taxonomic composition.

Transparent QC

Sample-wise and cohort-level quality metrics incorporated into project reporting.

Flexible Study Design

Support for individual samples, cohorts and comparative studies, with optional genome-resolved analysis.

Secure Data Delivery

Raw, intermediate and final datasets delivered through secure project storage.

Questions

Frequently asked, in plain language

From the basics to the finer points of design and statistics.

basicWhat is shotgun metagenomics?
It sequences random fragments of all DNA in a sample, so a single experiment shows which bacteria, archaea, fungi and viruses are present and which genes the community carries.
basicHow is it different from 16S sequencing?
16S sequencing reads one marker gene from bacteria and archaea. Shotgun sequencing reads everything, so it covers more kinds of organisms, gives species-level and functional information, and can recover genomes, at a higher cost per sample.
basicWhat will I receive?
Raw and processed FASTQ files, QC reports, taxonomic abundance tables from phylum to species, gene-family and pathway tables, assembled contigs and MAGs where included, diversity summaries, a summary report, and secure project storage access.
intermediateCan you analyse data I already have?
Yes. Pre-generated FASTQ files can be analysed without sequencing. Please share the read length, platform and how the DNA was extracted.
intermediateHow many reads per sample?
It depends on the sample type and the question. Simple communities need less, highly diverse ones such as soil need more, and host-rich samples need more again because host reads are discarded. Our team recommends a depth for your design.
intermediateWhy is host removal reported?
The share of host reads tells you how much of your sequencing yielded microbial data. A low microbial yield limits what can be detected, so it is shown for each sample.
intermediateIs a species-level result always certain?
No. Species calls depend on the reference database and on how well represented the genome is. Closely related species can be confused, and organisms missing from the database are assigned to relatives or left unclassified.
advancedAre the abundances absolute?
No. Sequencing gives proportions, so a rise in one taxon lowers the share of others. For absolute amounts, spike-ins or measurements of total microbial load are needed (Figure 9).
advancedWhen is assembly worthwhile?
When you want genomes, novel organisms or gene context rather than just abundance. It needs adequate depth and works best with several related samples. Recovered genomes are graded by completeness and contamination.
advancedWhich test finds differentially abundant taxa?
There is no single best test; methods disagree on real data. Results are most reliable when supported by more than one approach and when confounders such as batch are modelled. Share your design and covariates so the method can be chosen.
advancedDo you need negative controls?
For low-biomass samples, yes. Extraction blanks reveal contaminant taxa that otherwise look like real findings. We recommend including them in the submission.
References

Key literature & resources

Selected sources behind the methods and reference values on this page. Figures labelled "simulated", "schematic" or "illustrative" are generated for this page, not client data.

  1. Quince C, Walker AW, Simpson JT, Loman NJ, Segata N. Shotgun metagenomics, from sampling to analysis. Nature Biotechnology 2017;35:833–844.
  2. Sczyrba A, et al. Critical Assessment of Metagenome Interpretation: a benchmark of metagenomics software. Nature Methods 2017;14:1063–1071.
  3. Meyer F, et al. Critical Assessment of Metagenome Interpretation: the second round of challenges. Nature Methods 2022;19:429–440. Link
  4. Wood DE, Salzberg SL. Kraken: ultrafast metagenomic sequence classification using exact alignments. Genome Biology 2014;15:R46.
  5. Wood DE, Lu J, Langmead B. Improved metagenomic analysis with Kraken 2. Genome Biology 2019;20:257.
  6. Lu J, Breitwieser FP, Thielen P, Salzberg SL. Bracken: estimating species abundance in metagenomics data. PeerJ Computer Science 2017;3:e104.
  7. Blanco-Míguez A, et al. Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4. Nature Biotechnology 2023;41:1633–1644.
  8. Beghini F, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. eLife 2021;10:e65088.
  9. Li D, et al. MEGAHIT: an ultra-fast single-node solution for large and complex metagenomics assembly via succinct de Bruijn graph. Bioinformatics 2015;31:1674–1676.
  10. Nurk S, Meleshko D, Korobeynikov A, Pevzner PA. metaSPAdes: a new versatile metagenomic assembler. Genome Research 2017;27:824–834.
  11. Kang DD, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ 2019;7:e7359.
  12. Nissen JN, et al. Improved metagenome binning and assembly using deep variational autoencoders. Nature Biotechnology 2021;39:555–560.
  13. Sieber CMK, et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy (DAS Tool). Nature Microbiology 2018;3:836–843.
  14. Parks DH, et al. Recovery of nearly 8,000 metagenome-assembled genomes substantially expands the tree of life. Nature Microbiology 2017;2:1533–1542.
  15. Bowers RM, et al. Minimum information about a single amplified genome (MISAG) and a metagenome-assembled genome (MIMAG) of bacteria and archaea. Nature Biotechnology 2017;35:725–731.
  16. Chklovski A, Parks DH, Woodcroft BJ, Tyson GW. CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning. Nature Methods 2023;20:1203–1212. Preprint
  17. Chaumeil PA, et al. GTDB-Tk v2: memory friendly classification with the Genome Taxonomy Database. Bioinformatics 2022;38:5315–5316.
  18. Pasolli E, et al. Extensive unexplored human microbiome diversity revealed by over 150,000 genomes from metagenomes spanning age, geography, and lifestyle. Cell 2019;176:649–662.
  19. Almeida A, et al. A unified catalog of 204,938 reference genomes from the human gut microbiome. Nature Biotechnology 2021;39:105–114.
  20. Gloor GB, Macklaim JM, Pawlowsky-Glahn V, Egozcue JJ. Microbiome datasets are compositional: and this is not optional. Frontiers in Microbiology 2017;8:2224.
  21. Nearing JT, et al. Microbiome differential abundance methods produce different results across 38 datasets. Nature Communications 2022;13:342. Correction record
  22. Lin H, Peddada SD. Analysis of compositions of microbiomes with bias correction. Nature Communications 2020;11:3514.
  23. Vandeputte D, et al. Quantitative microbiome profiling links gut community variation to microbial load. Nature 2017;551:507–511.
  24. Salter SJ, et al. Reagent and laboratory contamination can critically impact sequence-based microbiome analyses. BMC Biology 2014;12:87.
  25. Eisenhofer R, et al. Contamination in low microbial biomass microbiome studies: issues and recommendations. Trends in Microbiology 2019;27:105–117.
  26. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 2014;30:2114–2120.
  27. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nature Methods 2012;9:357–359.
  28. Anderson MJ. A new method for non-parametric multivariate analysis of variance. Austral Ecology 2001;26:32–46.
  29. Aitchison J. The statistical analysis of compositional data. Journal of the Royal Statistical Society B 1982;44:139–177.
  30. Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society B 1995;57:289–300.
  31. Camargo AP, et al. Identification of mobile genetic elements with geNomad. Nature Biotechnology 2024;42:1303–1312.
  32. Shannon CE. A mathematical theory of communication. Bell System Technical Journal 1948;27:379–423.