July 22, 2024

cfDNA vs Genomic DNA mNGS: Which Gives More Signal?

Cell-free DNA (cfDNA) and genomic DNA (gDNA) can both be used for infectious-disease metagenomic next-generation sequencing (mNGS), but they do not answer the same sample-preparation problem. cfDNA workflows analyze short DNA fragments circulating in plasma or other body fluids. gDNA workflows analyze intact microbial genomes or longer microbial DNA fragments recovered from cells, pellets, or host-depleted sample fractions.

For laboratories comparing mNGS assay designs, the practical question is not “Which DNA type is always better?” It is: which workflow gives enough microbial signal, with manageable contamination risk and a sample type that fits the clinical question?

First-Screen Decision Matrix: cfDNA vs gDNA vs Host-Depleted gDNA

Choose cfDNA mNGS for plasma, CSF, or other acellular fluids where circulating microbial fragments are the expected diagnostic signal. The main limitation is that microbial DNA can be fragmented and close to background contamination.
Choose standard gDNA mNGS for tissue, pellets, swabs, or fluids with recoverable organisms and manageable host background. If host cells dominate, human DNA can consume sequencing depth before the sequencer captures microbial reads.
Choose host-depleted gDNA mNGS for whole blood, low-biomass fluids, culture-negative infection workups, and samples where intact organisms may be present but human DNA is high. Review the Devin Host Depletion Filter and the PaRTI-Seq workflow when host background is the bottleneck.

In the peer-reviewed sepsis study, host-depleted gDNA produced 9,351 microbial reads per million compared with 925 RPM for unfiltered gDNA. That is the practical click reason on this page: if the specimen contains intact organisms but host cells dominate, sample preparation can change the amount of microbial signal available before sequencing begins.

For cfDNA mNGS, the typical input is plasma, serum, CSF, or another acellular fluid. The main strength is noninvasive access to circulating microbial fragments. The main limitation is that low microbial abundance and fragmented genome coverage can complicate interpretation.

For gDNA mNGS, the typical input is whole blood, a cell pellet, tissue, a swab, or a host-depleted fraction. The main strength is more complete microbial sequence recovery when intact organisms are present. The main limitation is that the workflow needs sample preparation that reduces host-cell background without losing microbes.

The strongest infectious-disease use cases usually come from matching the DNA source to the specimen. cfDNA can be valuable when circulating microbial fragments are the relevant signal. Host-depleted gDNA becomes important when the sample contains intact organisms but human cells dominate the DNA background.

What cfDNA and Genomic DNA Mean in Infectious-Disease mNGS

Cell-free DNA is extracellular DNA released into blood or other fluids through cell death, tissue damage, pathogen turnover, or microbial shedding. In infectious-disease testing, the microbial fraction is often called microbial cell-free DNA (mcfDNA). It can include DNA from bacteria, fungi, parasites, and DNA viruses, but it is usually present at very low abundance relative to human cfDNA.

Genomic DNA in mNGS refers to DNA extracted from organisms or cells present in the specimen. In a microbial gDNA workflow, the goal is to recover DNA from bacteria, fungi, DNA viruses, or parasites while reducing the overwhelming host background. This is why sample preparation matters: without host depletion or microbial enrichment, human DNA can consume most sequencing reads.

The distinction matters most in blood. Plasma cfDNA workflows intentionally remove cells before extraction. gDNA workflows can instead use the cellular or microbial fraction, then apply host-cell depletion before extraction to enrich microbial DNA.

When cfDNA mNGS Works Well

cfDNA-based mNGS has clear advantages in selected infectious-disease contexts. A blood draw is less invasive than many tissue sampling procedures. Plasma cfDNA can be useful when pathogens shed nucleic acid into circulation, when culture is negative after antibiotics, or when clinicians need broad pathogen screening without committing to a narrow PCR panel.

Commercial cfDNA approaches have also built substantial clinical evidence. Karius now positions its plasma microbial cfDNA offering as Karius Spectrum, with clinical publications across bloodstream infection, pneumonia, invasive fungal infection, endocarditis, fever of unknown origin, and immunocompromised patient populations. Delve Bio’s Delve Detect focuses on cerebrospinal fluid, detecting bacteria, viruses, fungi, and parasites from 1 mL of CSF with reported 48-hour turnaround from sample receipt.

Those examples show why cfDNA remains important. For acellular fluids or situations where the microbial signal is expected to circulate as fragments, cfDNA can be a practical route to broad molecular detection.

Why cfDNA Can Miss Low-Abundance Microbial Signal

cfDNA also has constraints that matter for mNGS assay design.

First, the microbial fraction can be extremely small. Human cfDNA, background microbial DNA, reagent contaminants, and environmental DNA can all compete with the signal of interest. When pathogen-derived fragments are near the background level, distinguishing infection from contamination or transient DNAemia becomes harder.

Second, cfDNA is fragmented. Short, nonuniform genome coverage may be enough for species-level detection, but it can limit strain resolution, antimicrobial resistance gene context, and other genomic traits that depend on broader genome recovery.

Third, pre-extraction host-cell depletion cannot enrich cfDNA in the same way it can enrich cellular or microbial fractions. Once the workflow starts from plasma, nucleated host cells have already been removed, and the remaining problem is low microbial DNA abundance mixed with host cfDNA.

These limitations do not make cfDNA unsuitable. They mean cfDNA results need careful interpretation, especially for low-biomass samples and organisms that are not actively shedding detectable DNA into the tested fluid.

How Host-Depleted gDNA Changes the Sample-Prep Equation

Host-depleted gDNA workflows approach the blood problem differently. Instead of asking whether enough microbial fragments are already circulating in plasma, the workflow tries to retain intact microbes while removing host nucleated cells before DNA extraction.

Micronbrane’s Devin Host Depletion Filters are designed for this step. The workflow removes host cells from fluid specimens while allowing microorganisms to pass through, increasing the fraction of sequencing reads available for microbial detection. The resulting DNA can then be processed with Unison Low-Input Library Preparation Kits and analyzed through PaRTI-Seq bioinformatics.

This matters because sequencing capacity is finite. If most reads are human, the assay spends sequencing budget on material that does not identify the pathogen. A host-depleted gDNA workflow aims to shift read allocation toward microbial DNA before library preparation begins.

Evidence from the Published Host-Depletion Study

The key evidence for this comparison now comes from the peer-reviewed 2025 Molecular Diagnosis & Therapy study, “Optimization of Metagenomic Next-Generation Sequencing Workflow with a Novel Host Depletion Method for Enhanced Pathogen Detection.” The study evaluated a ZISC-based filtration device for host-cell depletion in blood mNGS and compared filtered gDNA, unfiltered gDNA, and cfDNA workflows in blood culture-positive sepsis samples.

In the clinical comparison, filtered gDNA detected the expected pathogens in all 8 evaluated blood culture-positive samples. Average microbial reads increased from 925 reads per million (RPM) without filtration to 9,351 RPM with filtration, a greater than tenfold increase. The same paper reports that cfDNA-based mNGS showed inconsistent sensitivity and was not significantly enhanced by filtration, with cfDNA microbial RPM in the 1,251 to 1,488 range.

Those numbers update the older preprint framing on this page. The practical conclusion is narrower and stronger: in blood samples where intact organisms are present, host depletion can materially improve the microbial signal available to a gDNA-based mNGS assay.

For more context on the same study, see Micronbrane’s summary: Peer-Reviewed Study Demonstrates Improved Sensitivity of mNGS Using Host Depletion Filtration.

Practical Workflow-Selection Guidance

Choose cfDNA-based mNGS when the specimen and clinical question fit an acellular-fragment workflow. Plasma and CSF cfDNA approaches can be useful when broad screening is needed, when an invasive biopsy is not practical, or when circulating microbial DNA is the expected diagnostic signal.

Choose host-depleted gDNA mNGS when the specimen may contain intact organisms but host DNA is expected to dominate the sequencing library. Whole blood, cell-rich fluids, and low-biomass samples are the most relevant cases. In those settings, microbial enrichment before extraction can improve read allocation and reduce the chance that the assay spends most of its sequencing depth on human DNA.

Use controls and contamination-aware interpretation in both workflows. Low-biomass mNGS is vulnerable to reagent and environmental DNA regardless of whether the input is cfDNA or gDNA. Negative controls, microbial DNA-free reagents, and consistent reporting thresholds are essential for both assay designs.

Micronbrane Workflow for Host-Depleted gDNA mNGS

Micronbrane’s PaRTI-Seq workflow is built around the gDNA side of this decision. It combines Devin host depletion, microbial DNA enrichment, low-input library preparation, and bioinformatics analysis to help laboratories recover more microbial signal from difficult infectious-disease specimens.

The workflow is most relevant when a lab wants to evaluate whether host-cell depletion improves microbial read yield before sequencing. Start with the mNGS workflow overview, then review the Devin Host Depletion Filter and Unison Library Preparation pages for sample-preparation details.

Contact Micronbrane to discuss whether host-depleted gDNA mNGS fits your sample type and study design.

References

Han D, Li R, Shi J, Tan P, Zhang R, Li J. Liquid biopsy for infectious diseases: a focus on microbial cell-free DNA sequencing. Theranostics. 2020;10(12):5501-5513. doi:10.7150/thno.45554.
Mandel P, Metais P. Les acides nucleiques du plasma sanguin chez l’homme. C R Seances Soc Biol Fil. 1948;142:241-243.
Camargo JF, Ahmed AA, Lindner MS, et al. Next-generation sequencing of microbial cell-free DNA for rapid noninvasive diagnosis of infectious diseases in immunocompromised hosts. F1000Research. 2019;8:1194.
Chen YC, Liao PH, Chen YW, et al. Optimization of Metagenomic Next-Generation Sequencing Workflow with a Novel Host Depletion Method for Enhanced Pathogen Detection. Molecular Diagnosis & Therapy. 2025;29:689-699. doi:10.1007/s40291-025-00797-3.
Thompson RA-C, et al. Utility of microbial cell free DNA next-generation sequencing for diagnosis and management of infectious diseases. Diagnostic Microbiology and Infectious Disease. 2024;110(1):116334.
Miller S, et al. Laboratory validation of a clinical metagenomic sequencing assay for pathogen detection in cerebrospinal fluid. Genome Research. 2019;29(5):831-842.