Long-Read mNGS with Host Depletion and PCR: AGBT Poster Results
At the September 2024 AGBT Precision Health conference, Micronbrane Medical presented a poster demonstrating two complementary approaches to long-read metagenomic sequencing on Oxford Nanopore Technologies (ONT) platforms. The poster, titled “Enabling Long-Read Next-Generation Sequencing for Metagenomics Applications,” addresses two persistent challenges in nanopore-based mNGS: detecting low-abundance microbial species with limited DNA input and recovering microbial signal from host-dominated clinical samples.
Download the poster: AGBT-PH Poster — Enabling Long-Read NGS for Metagenomics (PDF)
Key Takeaways
- Host depletion using the Devin Microbial DNA Enrichment Kit reduced human reads from 97% to 14% and enriched microbial reads from 0.34% to 85.96% in spiked blood samples sequenced on ONT Flongle.
- An optimized 16S+ITS PCR workflow detected all 21 reference microorganisms from the ZymoBIOMICS D6331 standard, achieving better correlation with expected microbial percentages than whole-genome sequencing alone.
- Both WGS and 16S+ITS workflows confidently detected microbial copy numbers as low as 100-1000.
- The Rapid PCR Barcoding Kit rescued host-depleted samples with DNA concentrations below the Qubit detection limit, yielding 63 ng/ul library DNA sufficient for Flongle sequencing.
- The combined host depletion and PCR protocol reduces cost and complexity by enabling informative long-read mNGS on the affordable ONT Flongle platform.
WGS and 16S+ITS Workflows for Nanopore Metagenomics
The poster evaluated two library preparation strategies using the ZymoBIOMICS D6331 microbial community standard, which contains 21 reference microorganisms at defined ratios.
Whole-genome sequencing (WGS): Starting from 200 ng of D6331 genomic DNA, the team prepared libraries using ONT’s standard Rapid Barcoding Kit. This approach provides unbiased detection across the microbial genome but requires a relatively high DNA input — a threshold that clinical specimens rarely meet after host depletion.
16S+ITS PCR: Starting from just 5 ng of D6331 DNA, the team amplified full-length 16S ribosomal RNA gene targets for bacteria and ITS (internal transcribed spacer) targets for fungi using an in-house PCR protocol, then prepared libraries for ONT sequencing. This amplicon-based approach detected all reference organisms and produced microbial composition profiles that correlated more closely with expected percentages than the WGS workflow.
Full-length 16S sequencing on nanopore platforms is particularly valuable because the long reads span the entire gene, enabling species-level taxonomic classification that short-read 16S approaches — which target only one or two variable regions — cannot reliably achieve. The inclusion of ITS primers in the same workflow extends coverage to fungal pathogens, an important consideration for comprehensive metagenomics.
Both workflows detected copy numbers as low as 100-1000, confirming that ONT sequencing has sufficient sensitivity for metagenomics applications when paired with appropriate library preparation.
Host Depletion Transforms Microbial Read Yield
The second part of the poster addressed the central challenge of clinical mNGS: host DNA dominance. The team spiked 2.5 mL of human whole blood with 10^6 copies of the ZymoBIOMICS Spike-in Control (D6320), then processed matched aliquots with and without host depletion using the Devin Microbial DNA Enrichment Kit.
Without host depletion, 97% of sequencing reads mapped to the human genome, and only 0.34% were classified as microbial. After host depletion, human reads dropped to 14% and microbial reads rose to 85.96% — a roughly 250-fold enrichment of the microbial fraction.
The host-depleted sample presented a practical challenge: extracted DNA concentration fell below the Qubit detection limit (<0.1 ng/ul), making standard library preparation impossible. The team addressed this by using ONT's Rapid PCR Barcoding Kit, which amplified the depleted DNA to produce a library at 63 ng/ul — well above the minimum input for Flongle sequencing. Microbial species identified from the depleted sample included Allobacillus halotolerans and Imtechella halotolerans, both present in the spike-in control.
The PCR amplification step adds time to the workflow, but optimized sequencing parameters achieve substantial read output within the first 12 hours of a Flongle run, keeping the total turnaround practical for laboratory settings.
Practical Implications for Low-Input mNGS
The AGBT poster data support several conclusions relevant to laboratories considering long-read mNGS adoption.
First, ONT platforms are viable for both WGS-based and amplicon-based (16S+ITS) metagenomics workflows. The choice depends on the application: WGS provides genome-wide coverage for antimicrobial resistance profiling and strain typing, while 16S+ITS offers more consistent taxonomic composition with lower DNA input requirements.
Second, host depletion is essential for clinical samples. The Devin host depletion filter converts samples that would yield almost no microbial information into specimens where the majority of sequencing output is microbial. Without this step, the cost of generating equivalent microbial coverage on Flongle would be prohibitive.
Third, the Rapid PCR Barcoding Kit bridges the gap between host-depleted ultralow-input samples and the DNA mass requirements of ONT library preparation. This combination makes the affordable Flongle flow cell a practical option for mNGS, reducing both per-sample cost and instrument complexity compared to higher-throughput ONT platforms.
For detailed sequencing kinetics and protocol optimization data, see the companion blog post: Long-Read mNGS on Flongle: Host Depletion + PCR.
About the Authors
Maurice Chan, Yizhen Lin, Deng Fong Chao, Jiahao Chang, Tian-Hsiang Ma, Hannah Lucio, Hau Hung, and Mengchu Wu are researchers and engineers at Micronbrane Medical, headquartered in Hsinchu, Taiwan. The team develops host depletion technologies, library preparation kits, and bioinformatics tools for metagenomic next-generation sequencing. Mengchu Wu is the CEO of Micronbrane Medical.
