NGS Library Prep Kits for Metagenomics: Low-Input Selection Guide
Why mNGS Library Prep Differs from Standard NGS
New DNA library prep kits for metagenomics are moving away from general whole-genome sequencing assumptions and toward low-input, mixed-community samples. For mNGS, the useful question is whether the kit still builds representative libraries when DNA input falls below 10 nanograms, or even below Qubit detection after host depletion. The key developments are low-input chemistry, host-depleted sample compatibility, Illumina versus ONT workflow fit, and indexing strategies that still work when microbial DNA is scarce.
For microbiology labs, the choice is not just “which NGS library prep kit is most popular?” It is whether the kit fits the specimen, platform, and contamination-control workflow: CSF, vitreous humor, plasma, BAL, and host-depleted swabs often need ultralow-input library construction, while ONT workflows may require PCR barcoding and different adapter chemistry.
If you want a direct Illumina kit benchmark, start with NGS Library Prep Kit Comparison: Unison vs NEB vs Illumina. This guide has a different job: choosing the right kit type for a microbiology or metagenomic sequencing workflow across sample types, platforms, and host-depletion conditions.
How to Choose an NGS Library Prep Kit for Microbiology
For microbiology NGS library prep, choose the kit by specimen type, platform, DNA input, and contamination-control needs. A kit that works for cultured isolate WGS may fail when the same lab processes host-depleted CSF, vitreous humor, plasma, BAL, or swabs for metagenomic sequencing.
Use this order:
- Start with sample type. Low-volume or low-biomass specimens need a kit that still performs when microbial DNA is scarce.
- Check post-depletion DNA input. Host depletion improves microbial read share but can leave sub-nanogram total DNA.
- Select by platform. Illumina-compatible kits and Oxford Nanopore kits use different adapter chemistry and should not be judged by the same benchmark alone.
- Confirm indexing and multiplexing. Indexing options matter when microbial input is low because index hopping, barcode balance, and no-template controls affect interpretation.
- Account for analysis. A kit is only useful if the resulting reads can be interpreted with validated contamination controls and a compatible pipeline such as PaRTI-Seq Analysis.
For `ngs library prep kit microbiology`, `microbiology ngs library prep ont`, and `dna library prep kit for metagenomics` searches, the short answer is this: use a low-input Illumina-compatible kit such as Unison when the workflow is host-depleted short-read mNGS, and use ONT-specific PCR barcoding or validated long-read protocols when real-time nanopore sequencing is the requirement.
Selection Guide by Sample Type
Standard whole-genome sequencing (WGS) library preparation protocols are designed for abundant, high-purity DNA inputs — typically 50 to 200 nanograms from cultured isolates or tissue. These protocols work well when DNA is plentiful, fragmentation is controlled, and the target genome is uniform.
Metagenomic next-generation sequencing (mNGS) operates under fundamentally different conditions. Clinical specimens such as cerebrospinal fluid, vitreous humor, pediatric plasma, and host-depleted swabs contain mixed-community DNA at concentrations that frequently fall below 10 nanograms — and often below the detection limit of the Qubit dsDNA assay. The DNA is a heterogeneous mixture of microbial and residual host fragments at variable sizes and ratios.
When a library prep kit designed for high-input WGS encounters sub-nanogram mNGS inputs, three things typically go wrong. Adapter ligation efficiency drops, producing fewer sequencing-ready fragments. PCR amplification introduces compositional bias, enriching certain taxa while suppressing others. And replicate consistency collapses, making it impossible to distinguish biological signal from technical noise. For irreplaceable clinical specimens — samples that cannot be recollected — a failed library construction wastes reagents, loses sequencing capacity, and eliminates diagnostic opportunity.
Choosing the right library prep kit for mNGS is not about finding the most popular kit. It is about finding one that maintains taxonomic fidelity at the DNA inputs your samples actually produce.
Kit Comparison: Illumina and Oxford Nanopore Platforms
The list below compares six library preparation kits across the two major sequencing platforms used in mNGS. Performance data comes from controlled benchmarking studies using serial dilutions of stool-derived microbial DNA (Illumina kits) and simulated clinical samples spiked with bacteria (ONT kits).
- Unison Ultralow DNA NGS Library Prep Kit on Illumina: 10 pg minimum input and approximately 45 minutes of hands-on time. It maintained taxonomic fidelity to 1 ng and preserved phylum-level composition at 0.1 ng without amplification artifacts.
- NEBNext Ultra II on Illumina: approximately 100 pg minimum input and approximately 2 hours of hands-on time. It showed distortion below 1 ng and Actinobacteria enrichment at 0.1 ng, a hallmark of low-input amplification bias.
- Illumina DNA Prep on Illumina: approximately 1 ng minimum input and approximately 2.5 hours of hands-on time. It showed distortion at 0.1 ng, with amplification bias visible in phylum-level bar charts.
- Illumina TruSeq DNA Nano on Illumina: 100 ng minimum input and approximately 3 hours of hands-on time. It cannot be tested below 100 ng because of protocol input limitations.
- ONT Rapid Barcoding Kit on Oxford Nanopore: 200 ng minimum input and approximately 70 minutes of hands-on time. It produced approximately 70,000 total reads from non-depleted samples, which is insufficient for low-input specimens.
- ONT Rapid PCR Barcoding Kit on Oxford Nanopore: 1 ng minimum input and approximately 3.5 hours of hands-on time. It produced approximately 270,000 total reads from host-depleted samples, roughly 4x more than the standard kit.
The performance gap at low inputs is stark. At 1 nanogram, the Unison kit maintained consistent alpha diversity (Shannon and Simpson indices) across replicates, while competing Illumina kits showed significant distortion. On the ONT side, the PCR barcoding approach is the only viable option for host-depleted samples, where DNA concentration falls below Qubit detection.
For detailed benchmarking methodology and diversity plots, see Overcoming Library Prep Failure in Low-Biomass and Host-Depleted Samples.
The Low-Biomass Challenge: Why Most Kits Fail
The scale of the input gap in mNGS is often underappreciated. Consider the ONT Flongle flow cell: it has a theoretical output of approximately 2.6 Gb, translating to roughly 650,000 reads. The total mass of DNA that actually gets sequenced is approximately 2.8 picograms. Yet the Rapid Barcoding Kit requires 200 nanograms of input — a 70,000-fold excess over what the instrument needs.
This gap exists because library preparation protocols must account for ligation efficiency, adapter incorporation rates, and loading optimization. Most protocols were calibrated for WGS applications where DNA is abundant. In mNGS, especially after host depletion, the available DNA often falls below 1 nanogram — or below Qubit detection entirely.
When insufficient library DNA is loaded onto an ONT flow cell, the sequencing duty plot degrades: fewer pores are actively sequencing, output drops well below capacity, and coverage gaps appear across target genomes. On Illumina platforms, low-input libraries produce cluster densities too sparse for reliable base calling, and the compensatory PCR amplification introduces the compositional distortions described above.
The clinical sample types most affected include cerebrospinal fluid, vitreous humor, pediatric plasma, bronchoalveolar lavage, and host-depleted swabs. For laboratories working with these specimen types, library prep failure at low DNA inputs is not an edge case — it is the expected condition.
How Host Depletion Before Library Prep Improves Results
Host depletion is the upstream step that makes low-input library preparation necessary — and worthwhile. The Devin Host Depletion Filter removes approximately 99% of human nucleated cells from clinical samples in under five minutes, shifting the read composition from over 95% human to predominantly microbial.
This transformation has a direct impact on library prep outcomes. On the ONT platform, host-depleted samples processed with the Rapid PCR Barcoding Kit produced approximately 270,000 total reads — nearly 4x more than non-depleted samples processed with the standard Rapid Barcoding Kit. The microbial read percentage increased by orders of magnitude, enabling genome assembly and antimicrobial resistance gene prediction that would be impossible without depletion.
On the Illumina platform, host depletion reduces the sequencing depth required per sample. When 99% of human reads are removed before sequencing, laboratories can achieve diagnostic-quality microbial data at approximately 5 million reads per sample — a four-fold reduction compared to standard protocols. This shifts the economics of platform choice: even mid-output benchtop instruments become viable mNGS tools when paired with effective host depletion.
The pairing matters. Host depletion creates an ultralow-biomass DNA challenge that demands a library prep kit capable of working at sub-nanogram inputs. A kit that fails below 10 nanograms eliminates the advantage that host depletion provides. For a side-by-side comparison of charge-based filtration, enzymatic lysis, and other depletion methods — including cost-per-sample data — see XNA vs Enzymatic Host Depletion: Which Method Fits Your Workflow?. For a complete sample-to-sequencing workflow, see the PaRTI-Seq assay, which integrates host depletion, contamination-controlled extraction, ultralow library preparation, and bioinformatics analysis into a single protocol.
Short-Read vs Long-Read: Platform-Specific Kit Selection
The choice between Illumina and ONT platforms affects which library prep kits are available and how they perform in mNGS workflows.
Short-read platforms (Illumina) offer the widest selection of library prep kits, higher throughput, and lower per-read costs. Multiple competing kits serve different input ranges, and the sequencing-by-synthesis chemistry is well-established for metagenomic applications. The critical selection criterion for mNGS is not cost per gigabase — the standard WGS metric — but cost per 5 million reads, since that depth is typically sufficient for pathogen identification with an optimized workflow.
Long-read platforms (ONT) provide real-time sequencing, long reads that improve taxonomic classification and AMR gene context, and low capital costs. However, library prep options are more limited, and host-depleted samples almost always require PCR-based amplification. The ONT Rapid PCR Barcoding Kit with controlled PCR cycles is currently the only practical approach for sub-nanogram mNGS inputs on the platform.
The turnaround time advantage of ONT is significant for clinical applications. The optimized ONT library preparation protocol achieves approximately 6 hours from sample to sequencing — including 2 hours for host depletion and DNA extraction, plus 3.5 hours for library construction. Within 3 hours of sequencing, the PCR protocol surpasses the total read output of the standard protocol’s complete 24-hour run.
For a detailed platform-by-platform comparison of throughput, cost, and turnaround time across emerging sequencers, see Short Read Sequencing Platforms for mNGS: A Comparison.
Decision Framework: Which Kit for Which Application
The right library prep kit depends on two factors: your expected DNA input and your sequencing platform.
If your input DNA exceeds 100 ng (standard clinical isolates, non-depleted specimens): Any standard library prep kit for your platform will work. The TruSeq DNA Nano or Illumina DNA Prep are proven choices for Illumina; the Rapid Barcoding Kit works for ONT.
If your input DNA is 1–100 ng (partially depleted samples, moderate-biomass specimens): On Illumina, the Unison Ultralow DNA NGS Library Prep Kit or NEBNext Ultra II are both viable. Unison maintains tighter replicate consistency at the lower end of this range. On ONT, the Rapid PCR Barcoding Kit is recommended.
If your input DNA is below 1 ng (host-depleted clinical specimens, CSF, vitreous humor): On Illumina, only the Unison Ultralow DNA NGS Library Prep Kit has demonstrated maintained taxonomic fidelity at these inputs. It requires as little as 10 picograms and completes in 45 minutes of hands-on time. On ONT, the Rapid PCR Barcoding Kit with optimized cycle numbers is the only option.
If you are pairing host depletion with library prep: The Devin Host Depletion Filter paired with the Unison kit (Illumina) or Devin paired with the Rapid PCR Barcoding Kit (ONT) creates an end-to-end workflow for reducing mNGS costs while improving microbial reads. Host depletion removes 99% of human cells; the paired library prep kit recovers sequencing-ready libraries from the resulting trace DNA.
For host-depleted specimens that fall below standard DNA input thresholds, review the Unison Ultralow DNA NGS Library Prep Kit as the library construction step in the broader PaRTI-Seq workflow.
Microbiology FAQ
Which NGS library prep kit is used for microbiology? Microbiology NGS library prep depends on the sample type and DNA input. Cultured isolates with abundant DNA can use standard WGS library prep kits, but metagenomic microbiology samples usually contain mixed microbial DNA, residual host DNA, and very low total input. For those samples, the best-fit kit is one that maintains taxonomic fidelity below 1 ng and does not distort community composition during amplification.
What sample types work best with DNA library prep kits with indexing options? Indexed DNA library prep works best when the lab can balance input amount, barcode design, and contamination controls across the run. Cultured isolates and higher-biomass samples are straightforward. Low-biomass mNGS samples need more care: CSF, vitreous humor, plasma, BAL, and host-depleted swabs should use low-input chemistry, no-template library controls, and index layouts that reduce barcode imbalance or carryover risk.
Should microbiology labs use Illumina or Oxford Nanopore library prep? Use Illumina-compatible library prep when the goal is high-accuracy short reads, established mNGS workflows, and efficient multiplexing around roughly 5 million reads per sample. Use Oxford Nanopore library prep when real-time sequencing, long reads, rapid field deployment, or resistance-gene context are more important. ONT workflows usually need ONT-specific adapters and often PCR barcoding for low-input host-depleted samples.
What are popular NGS library prep options for metagenomics? Common options include Illumina DNA Prep, NEBNext Ultra II, TruSeq DNA Nano, Unison Ultralow DNA NGS Library Prep Kit for Illumina workflows, and ONT Rapid or Rapid PCR barcoding kits for nanopore workflows. The important metagenomics comparison is not popularity alone; it is whether the kit works after host depletion, supports low-biomass clinical specimens, and preserves the relative abundance of organisms in the sample.
When should a microbiology lab choose an ultralow-input kit? Choose an ultralow-input kit when the sample is host-depleted, difficult to recollect, or expected to produce sub-nanogram microbial DNA. This is common for CSF, vitreous humor, pediatric plasma, bronchoalveolar lavage, and swabs after host-cell removal. In those cases, a high-input WGS kit can fail library construction or create biased readouts even if the sequencer itself has enough capacity.
