Key takeaways
- 16S/ITS amplicon: the lowest cost per sample. Good for comparing community composition across many samples, usually at genus level.
- Full-length 16S (long reads): better species-level resolution, still amplicon-based.
- Shotgun metagenomics: all domains (bacteria, archaea, fungi, viruses), species or strain level, plus functional genes such as antibiotic resistance. Costs more per sample.
- For host-rich samples, amplicon methods are often the practical choice.
The three main options
16S / ITS amplicon sequencing
A PCR amplifies one marker gene from every microbe in the sample: a variable region of the 16S rRNA gene for bacteria and archaea (commonly V3–V4), or the ITS region for fungi. The amplicons are sequenced on a short-read instrument and grouped into taxa. Because only one gene is sequenced, a modest number of reads per sample is enough, and many samples share a run.
Full-length 16S
Long-read sequencing reads the entire ~1,500-base 16S gene instead of one region. The extra information separates closely related species that look identical on a short region.
Shotgun metagenomics
All the DNA in the sample is fragmented and sequenced, with no PCR target. Reads are matched to reference genomes or assembled into metagenome-assembled genomes (MAGs). You see every domain of life and the genes each organism carries.
| 16S / ITS amplicon | Full-length 16S | Shotgun metagenomics | |
|---|---|---|---|
| Organisms covered | Bacteria and archaea (16S) or fungi (ITS) | Bacteria and archaea | Bacteria, archaea, fungi, viruses, host |
| Typical resolution | Genus | Species | Species to strain |
| Functional genes | Predicted only | Predicted only | Measured directly (e.g. AMR genes, pathways) |
| Reads per sample | Tens of thousands (more for complex communities) | Thousands to tens of thousands of long reads | Millions or more (more for complex communities) |
| Host DNA | Not a problem | Not a problem | Can consume most of the reads |
| Cost per sample | Lowest | Moderate | Highest |
How to choose
- Large cohort, comparing groups (treatment vs control, time series): 16S amplicon gives the most samples for your budget.
- You need species names, for example to follow a specific probiotic or pathogen: full-length 16S or shotgun.
- You need function, such as resistance genes, metabolic pathways or virulence factors: shotgun metagenomics.
- Fungi matter: add ITS sequencing, or use shotgun.
- Viruses matter: shotgun (DNA viruses) or metatranscriptomics (RNA viruses).
- Tissue, biopsies or low-biomass samples: start with amplicon sequencing, or discuss host-depletion options before choosing shotgun.
A common and cost-effective design is 16S on every sample, then shotgun on a selected subset where the differences look interesting.
Sample collection matters as much as the method
- Use a consistent collection and storage method across the whole study (stabilisation buffer or immediate freezing).
- Include negative controls (extraction blanks) and, ideally, a mock community as a positive control. This is essential for low-biomass samples.
- Record metadata (time, site, treatment, storage) from the start; it drives the statistical analysis later.
Please send extracted, purified DNA. If you need help choosing an extraction method for your sample type, email or call our scientists.
Analysis
Amplicon data is typically processed into amplicon sequence variants (ASVs), assigned taxonomy, and analysed for diversity and differential abundance. Shotgun data adds taxonomic profiling at the species level, functional profiling and, where depth allows, genome assembly. All of these can be added as bioinformatics analysis, or you can receive raw FASTQ files for your own pipeline.
Frequently asked questions
Can 16S sequencing identify bacteria to species level?
Short-read 16S (for example the V3–V4 region) usually resolves bacteria reliably to genus level, with species calls only for some groups. Full-length 16S on long reads improves species-level resolution. Shotgun metagenomics can go to species and often strain level.
Does 16S sequencing detect fungi and viruses?
No. 16S targets bacteria and archaea. Fungi are profiled with ITS amplicon sequencing, and viruses need shotgun metagenomics (or metatranscriptomics for RNA viruses).
How many reads do I need per sample?
It depends mostly on how complex the community is: the more different microbes a sample contains, the deeper you need to sequence to detect the rare ones. As a rule of thumb, tens of thousands of reads per sample for 16S/ITS amplicon surveys, and several million or more for shotgun metagenomics. Diverse samples such as soil or stool, and host-rich samples, need more.
Which method is better for samples with a lot of host DNA?
16S amplicon sequencing, because the PCR targets only microbial DNA. In shotgun sequencing of tissue, biopsies or some swabs, most reads can come from the host, which raises the depth (and cost) needed to see the microbes.
Can I measure which genes are active?
Not with DNA methods. Shotgun metagenomics shows which genes are present; metatranscriptomics sequences RNA to show which genes are being expressed.
