Long-Read mNGS on Flongle: Host Depletion + PCR
Metagenomic next-generation sequencing (mNGS) holds enormous potential for unbiased pathogen detection, but two practical barriers have slowed its adoption in routine laboratory settings: the overwhelming presence of host DNA in clinical samples and the high cost of generating sufficient sequencing depth. Long-read sequencing platforms from Oxford Nanopore Technologies (ONT) offer a compelling path forward, yet their utility for mNGS depends on solving these upstream challenges first.
Recent application data demonstrate that pairing host depletion with PCR-based library amplification can unlock cost-effective long-read metagenomic sequencing, even on the most affordable ONT hardware available today.
The Challenge: Host DNA Dominance in mNGS
In a typical clinical blood sample, human nucleated cells vastly outnumber microbial cells. Without intervention, host-derived reads can account for over 97% of total sequencing output, leaving only a tiny fraction available for microbial detection. This imbalance forces laboratories into a difficult trade-off: either sequence at extreme depth to capture enough microbial signal (driving up cost and turnaround time) or accept reduced sensitivity.
As outlined in our earlier post on 4 ways to get more microbial reads for less cost, host depletion is the single most impactful step a laboratory can take to improve the economics of mNGS. Removing human cells before DNA extraction shifts the sequencing effort toward the organisms that matter most.
Why Long-Read Sequencing Changes the Game
Short-read platforms have dominated mNGS to date, but long-read sequencing offers distinct advantages for metagenomic applications. Reads spanning several kilobases improve taxonomic classification accuracy, resolve repetitive genomic regions, and enable more confident assembly of microbial genomes directly from complex samples. Long reads also simplify the detection of structural variants and antimicrobial resistance gene context, information that is difficult to reconstruct from short fragments alone.
ONT platforms are particularly attractive because of their real-time data streaming, portable form factors, and low capital cost. The Flongle flow cell, priced at approximately $90 per run, brings per-sample sequencing costs within reach of routine laboratory budgets. However, the Flongle’s lower throughput ceiling (approximately 2.6 Gb and 650,000 reads at a mean fragment length of 4 kb) means that every read counts. Wasting the majority of output on host DNA is not an option.
The Protocol: Host Depletion + PCR Amplification
A combined workflow addresses both the host DNA problem and the input limitation of ultralow-biomass samples in a straightforward two-stage approach.
Stage 1: Host Depletion. The Devin™ Microbial Enrichment Kit depletes human nucleated cells from whole blood using Micronbrane Medical’s patented ZISC (Zwitterionic Interface Ultra-Self-assemble Coating) technology. In controlled experiments using human blood spiked with ZymoBIOMICS spike-in controls, host depletion reduced the proportion of human reads from 97% down to approximately 14%, while microbial species (including Allobacillus halotolerans and Imtechella halotolerans) were enriched from a combined 0.34-0.36% to 85.96% of total reads.
Stage 2: PCR Amplification. Host-depleted samples often yield ultralow quantities of total DNA, which can fall below the input threshold of standard library preparation kits. The ONT Rapid PCR Barcoding Kit addresses this by amplifying the library to a sufficient mass for sequencing. In testing, 20 PCR cycles performed comparably to 24 cycles, indicating that moderate amplification is sufficient without introducing excessive bias.
The total workflow time with the PCR amplification step is approximately 260 minutes from sample to loaded flow cell, compared to roughly 70 minutes for a non-PCR library preparation. While the PCR path is longer, it is the practical requirement for host-depleted samples where DNA yield is too low for direct library construction.
Key Results: From 97% Human Reads to 86% Microbial Reads
The performance gains from combining host depletion with PCR amplification are substantial:
- Read yield: PCR-amplified libraries produced 277,910 QC-passed reads and 1.06 Gb of data on a single Flongle flow cell, compared to 66,747-89,373 reads and 0.25-0.27 Gb from non-PCR libraries.
- Sequencing speed: Throughput increased from approximately 8,500 reads per hour to approximately 22,500 reads per hour, making better use of the flow cell’s active lifetime.
- Microbial enrichment: The Devin™ Host Depletion filter shifted the composition from 97% human reads to 86% microbial reads, a transformation that directly translates to more actionable data per sequencing dollar.
- Cost efficiency: Achieving over 1 Gb of data on a $90 Flongle flow cell means that laboratories can perform informative long-read metagenomic runs without committing to higher-cost MinION or PromethION flow cells.
These results demonstrate that the combination of host depletion and PCR amplification does not merely make long-read mNGS possible on budget hardware. It makes it practical.
Rapid Barcoding Kit vs. Rapid PCR Barcoding Kit: A Closer Look
The choice between ONT’s standard Rapid Barcoding Kit and the Rapid PCR Barcoding Kit is not simply a matter of preference — it is dictated by the DNA input available after host depletion. Non-depleted samples typically yield enough DNA for the standard Rapid Barcoding Kit (minimum 200 ng input). However, host-depleted clinical specimens frequently produce DNA concentrations below the Qubit detection limit, making direct library construction with the standard kit impossible.
Duty plot analysis from head-to-head testing revealed a counterintuitive finding: despite loading sufficient DNA per the manufacturer’s specifications, the Rapid Barcoding Kit achieved suboptimal sequencing efficiency. Fewer nanopores were actively sequencing at any given time, and the flow cell underperformed relative to its capacity. The Rapid PCR Barcoding Kit, by contrast, generated a visibly healthier duty plot with more pores engaged in active sequencing throughout the run.
The read output differences were substantial. PCR-amplified host-depleted libraries produced approximately 270,000 total reads, compared to roughly 70,000 reads from the non-PCR protocol with non-depleted samples — nearly a fourfold difference. While the non-PCR protocol generated longer median read lengths (consistent with the absence of amplification artifacts), the PCR protocol produced reads with a median length of approximately 4,000 base pairs, which remains more than sufficient for confident taxonomic classification and genome assembly.
The turnaround time tradeoff also deserves consideration. The PCR amplification step adds approximately three hours to library construction (total sample-to-sequencing time of roughly six hours versus 70 minutes for non-PCR). However, the PCR protocol’s higher sequencing efficiency compensates quickly: within three hours of sequencing, the PCR workflow surpasses the total read output that the non-PCR protocol generates across its entire 24-hour run. For laboratories that need results within a working day, the PCR path delivers more actionable data faster despite the longer preparation time.
For detailed protocol specifications and sequencing kinetics data, see the companion webinar summary: Optimized ONT Library Prep Protocol for Low-Input mNGS.
What This Means for Your Lab
For laboratories evaluating long-read mNGS, this workflow removes several adoption barriers at once. The Flongle’s low per-run cost reduces the financial risk of exploratory sequencing. Host depletion with the Devin™ Microbial Enrichment Kit ensures that the limited throughput is directed toward microbial content rather than wasted on human reads. And PCR amplification rescues ultralow-input samples that would otherwise fail library preparation entirely.
This approach is well suited for surveillance studies, point-of-care investigations in resource-limited settings, and any application where rapid, affordable, long-read metagenomic data would inform decision-making. The protocol is also compatible with ONT barcoding, enabling multiplexed runs that further reduce per-sample costs.
Get Started with the Devin Microbial Enrichment Kit
The Devin™ Microbial Enrichment Kit provides everything needed for the host depletion stage of this workflow, including mNGS-grade reagents designed to minimize background contamination. Whether you are running long-read or short-read platforms, host depletion is the foundation of cost-effective metagenomic sequencing.
To learn more about integrating the Devin kit into your long-read mNGS workflow, contact us at info@micronbrane.com.
