Skip to main content
Download PDF

Comparative analysis of metagenomic next-generation sequencing for pathogenic identification in clinical body fluid samples

BMC Microbiology volume 25, Article number: 165 (2025) Cite this article

Abstract

Objectives

This study aims to evaluate and compare the effectiveness of metagenomic next-generation sequencing (mNGS) in identifying pathogens from clinical body fluid samples, with a specific focus on the application of microbial cell-free DNA (cfDNA) mNGS.

Methods

A total of 125 clinical body fluid samples were collected. All samples underwent mNGS targeting whole-cell DNA (wcDNA), with 30 samples also analyzed for cfDNA mNGS and 41 subjected to 16S rRNA NGS for comparative analysis. Patient clinical data, including culture results, were obtained from electronic medical records.

Results

In comparison to cfDNA mNGS, the mean proportion of host DNA in wcDNA mNGS was 84%, significantly lower than the 95% observed in cfDNA mNGS (p < 0.05). Using culture results as a reference, concordance rates were 63.33% (19/30) for wcDNA mNGS and 46.67% (14/30) for cfDNA mNGS. Additionally, wcDNA mNGS showed greater consistency in bacterial detection with culture results, achieving a rate of 70.7% (29/41) compared to 58.54% (24/41) for 16S rRNA NGS. The sensitivity and specificity of wcDNA mNGS for pathogen detection in body fluid samples were 74.07% and 56.34%, respectively, when compared to culture results.

Conclusion

Whole-cell DNA mNGS demonstrates significantly higher sensitivity for pathogen detection and identification compared to both cfDNA mNGS and 16S rRNA NGS in clinical body fluid samples, particularly those associated with abdominal infections. However, the compromised specificity of wcDNA mNGS highlights the necessity for careful interpretation in clinical practice.

Peer Review reports

Introduction

Culture is a routine method for the detection and identification of pathogens in clinical microbiology laboratories. While culture serves as an enrichment technique for microbial pathogens, it is not effective for identifying anaerobic, low-abundance, fastidious, and/or slow-growing microorganisms [1]. Additionally, with the rising prevalence of polymicrobial infections, traditional culture methods are often inadequate for identifying multiple pathogens simultaneously [2,3,4]. Although observing colony morphology aids in isolating and subculturing microbial pathogens, this approach may overlook organisms lacking distinct morphological features [5, 6]. Consequently, alternative methods, such as serological tests and nucleic acid amplification tests (NAATs), are frequently employed to enhance the accurate identification of pathogens in clinical settings [1, 7,8,9]. However, these methods often fail to effectively capture the diversity of pathogens present in clinical body fluid samples.

Currently, two culture-independent methods based on next-generation sequencing (NGS) are widely utilized for the detection and identification of microbial pathogens: 16S rRNA gene NGS (16S rRNA NGS) and metagenomic NGS (mNGS) [1, 8]. The 16S rRNA NGS approach typically involves amplification of the 16S rRNA gene [10, 11]. Although this method is prevalent for pathogen identification, it may not reliably differentiate some pathogens due to the high conservation of the 16S rRNA gene across many species [12]. In contrast, mNGS offers an unbiased, genome-wide analysis for the detection of both known and unknown pathogens [7, 9, 10, 13,14,15,16,17,18]. However, several studies indicate that excessive background or host DNA may negatively impact its sensitivity [7, 11, 15, 18]. Furthermore, it remains undetermined whether microbial cell-free DNA (cfDNA) serves as a superior target for NGS compared to whole-cell DNA (wcDNA). Most investigations primarily focus on cfDNA mNGS, with limited comparisons made regarding the clinical performances of cfDNA mNGS and wcDNA mNGS in pathogen detection [11, 13, 14, 16, 18,19,20].

A sterile body fluid sample is defined as a clinical specimen devoid of microbes, collected from a sterile site, with various sample types being highly diverse. In this study, we collected 125 clinical body fluid samples, which included pleural, pancreatic, drainage, ascites, and cerebrospinal fluid (CSF). We assessed the clinical performance of mNGS through a comparative analysis involving both cfDNA and wcDNA mNGS, as well as a comparison with 16S rRNA NGS.

Materials and methods

Study overview

This retrospective single-center study was approved by the Ethics Committee of Jinling Hospital in China (approval number: 2021DZGZR-YBB-013). We collected a total of 125 clinical body fluid samples (Fig. 1) submitted to the routine clinical microbiology laboratory at Jinling Hospital in Nanjing, stored at − 80 °C from December 2021 to February 2023. Patient and sample characteristics, including culture results, were retrieved from the hospital’s electronic medical records. Patients were excluded if they declined participation or if there were insufficient samples for DNA extraction. Written informed consent was obtained from all participants or from their legal guardians. We conducted three comparative analyses to assess the efficacy of mNGS in detecting and identifying pathogens in clinical body fluid samples (Fig. 1). Specifically, we evaluated 30 samples to assess the impact of host DNA on detection using both cfDNA mNGS and wcDNA mNGS methods. Additionally, 41 samples were analyzed using both 16S rRNA NGS and wcDNA mNGS. Overall, a total of 125 samples were assessed using wcDNA mNGS, with culture results serving as the benchmark for comparison.

Fig. 1
figure 1

Study overview and sample type composition. A total of 125 clinical body fluid samples were collected. Among these, thirty samples were used to compare cell-free DNA metagenomic next-generation sequencing (cfDNA mNGS) with whole-cell DNA metagenomic next-generation sequencing (wcDNA mNGS); forty-one samples were used to compare 16S rRNA gene NGS (16S rRNA NGS) with wcDNA mNGS; and all samples were used to evaluate the clinical performance of wcDNA mNGS was evaluated against culture results as a reference

Full size image

Sample processing and DNA extraction

Thirty clinical body fluid samples were centrifuged at 20,000 × g for 15 min, as previously described [7, 18]. Cell-free DNA was extracted from 400 μl of supernatant using the VAHTS Free-Circulating DNA Maxi Kit (Vazyme Biotech Co., Ltd., Nanjing, China), following the manufacturer’s instructions. Briefly, 25 μl of Proteinase K, 800 μl of Buffer L/B, and 15 μl of magnetic beads were added to the sample, mixed briefly, and incubated at room temperature for 5 min. The tube was then placed on a magnetic rack to allow the solution to clear, after which the supernatant was carefully removed. The sample was washed, and finally, 50 μl of elution buffer was added to resuspend the magnetic beads before transferring the supernatant into a new centrifuge tube. To extract wcDNA, two 3-mm nickel beads were added to the retained precipitate, which was then shaken at 3,000 rpm for 5 min to facilitate cell lysis. Subsequently, wcDNA was extracted from the precipitate using the Qiagen DNA Mini Kit (Qiagen, Shanghai, China) according to the manufacturer’s protocol and our previous study [21].

Metagenomic NGS

DNA library preparation was performed using the VAHTS Universal Pro DNA Library Prep Kit for Illumina (Vazyme Biotech Co., Ltd., Nanjing, China) following the manufacturer’s instructions, as described in our previous study [22]. NGS was conducted using the NovaSeq platform (Illumina, San Diego, CA, USA) with a 2 × 150 paired-end configuration. Each sample generated approximately 8 GB of sequencing data, corresponding to roughly 26.7 million reads. A total of 100 samples were tested in each run.

16S rRNA NGS

The library for 16S rRNA NGS was prepared as outlined in our previous study [21]. NGS was performed using the NovaSeq platform (Illumina, San Diego, CA, USA) with a 2 × 250 paired-end configuration. Each sample generated approximately 0.05 million reads.

Bioinformatic analysis

Data obtained from mNGS and 16S rRNA NGS were analyzed following protocols reported in previous studies [21, 22]. When operational taxonomic units (OTUs) from 16S rRNA NGS could not be accurately identified at the species level, species identification was performed by manually aligning the 16S rRNA sequences with known species on the NCBI website (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

Criteria for reporting pathogens

The criteria for clinical reporting of pathogens using 16S rRNA NGS or mNGS were developed based on prior studies [7, 9, 15, 18,19,20, 23, 24]. The percentage of read counts and z-scores for each species per sample were calculated using Pavian (https://fbreitwieser.shinyapps.io/pavian/) and compared with those in negative controls [25]. The criteria for identifying valid reportable pathogens using 16S rRNA NGS were: 1) a z-score of the species being threefold that of the negative control; 2) read counts exceeding 100; 3) a sample with no 16S rRNA gene amplification product was considered negative; 4) when reads mapped to multiple species within the same genus, the species with the highest read count was retained only if its read count was at least ten-fold greater than that of any other species. For mNGS, the criteria included: 1) a species-to-negative control z-score ratio greater than three; 2) reads that mapped to five different genomic regions; 3) read counts for bacteria greater than 100; 4) when reads were annotated to multiple species within the same genus, the species with the highest read count was selected only if its read count was at least five-fold greater than that of any other species; and 5) read counts for fungi or viruses greater than 10. Additionally, all pathogen detections considered contaminants, colonizers, and commensals were excluded from reporting.

Statistical analysis

All statistical analyses were performed using R software (version 4.3.3). Continuous variables were expressed as medians with ranges or as means with standard deviation, while categorical variables were represented as counts and percentages. The Student’s t-test was used to compare differences in host DNA proportions between cfDNA mNGS and wcDNA mNGS. Based on the consistency with culture results, we analyzed the differences between wcDNA mNGS and cfDNA mNGS, as well as between wcDNA mNGS and 16S rRNA NGS results using cross-tabulation. Additionally, Pearson correlation was employed to evaluate the consistency of pathogenic reads detected by wcDNA mNGS compared to cfDNA mNGS and 16S rRNA NGS. A chi-square test or Fisher's exact test was utilized to assess the clinical performance of wcDNA mNGS in comparison to culture. The significance level was set at a p-value of less than 0.05.

Results

Patients and samples

A total of 125 clinical body fluid samples were collected from 107 hospitalized patients, as depicted in Fig. 1. All samples were cultured and underwent wcDNA mNGS. Of these samples, 30 samples were subjected to cfDNA mNGS, and 41 samples underwent 16S rRNA NGS, allowing for comparative analysis with the results of wcDNA mNGS. Patients had a mean age of 50 years, with 69% being male. The median duration of hospitalization was 32 days, and the median levels of C-reactive protein and procalcitonin were 65 mg/L and 0.29 µg/L, respectively. The sample types included: ascites (n = 24), bile (n = 27), CSF (n = 16), exudate (n = 1), gallbladder drainage fluid (n = 1), gastric fluid (n = 1), joint fluid (n = 5), nephropuncture fluid (n = 1), pancreatic fluid (n = 18), pericardial fluid (n = 1), peritoneal drainage fluid (n = 7), pleural fluid (n = 17), renal drainage fluid (n = 3), and tissue (n = 3). Clinical signs and symptoms in each patient indicated the presence of infection, leading to microbiological testing post-surgery.

Comparison sequencing read counts between cfDNA and wcDNA mNGS

We analyzed 30 body fluid samples to compare the effects of cfDNA and wcDNA on pathogen detection using mNGS. The mean read counts from cfDNA mNGS and wcDNA mNGS were 2.4 × 10^7 and 2.7 × 10^7, respectively. Raw read counts from cfDNA mNGS were significantly lower than those from wcDNA mNGS (p < 0.05; Fig. 2A), although no significant differences were observed among different sample types (Figure S1).

Fig. 2
figure 2

Analysis of paired sequencing data obtained using cfDNA mNGS and wcDNA mNGS. A Raw read counts from cfDNA mNGS and wcDNA mNGS; B-C Read counts and Proportions of host DNA; D Classified reads obtained from cfDNA mNGS and wcDNA mNGS using Kraken2 for annotation

Full size image

For host DNA, wcDNA mNGS yielded higher mean read counts than cfDNA mNGS (Fig. 2B and Figure S2), except for bile samples. The mean proportion of host DNA was 84% for wcDNA mNGS, lower than the 95% seen in cfDNA mNGS (Student’s t-test, p < 0.05, Fig. 2C). Notably, the proportion of host DNA varied by sample type; for example, CSF samples showed a proportion of 98% for cfDNA and nearly 100% for wcDNA (Figure S3). Thus, while host DNA read counts were higher in abdominal infection samples compared to CSF, their proportion was relatively low. Overall, the mean classified read counts of cfDNA mNGS were slightly higher than those of wcDNA mNGS, though the difference was not statistically significant (Fig. 2D). Additionally, differences in classified read counts varied by clinical samples; for instance, cfDNA mNGS classified read counts were higher in CSF but lower in pancreatic fluid compared to wcDNA mNGS (Figure S4).

Comparison of pathogen identification using cfDNA and wcDNA mNGS relative to culture

In evaluating 30 body fluid samples, we compared the clinical performance of cfDNA mNGS and wcDNA mNGS against culture results (Table S1). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for cfDNA mNGS were all lower than those for wcDNA mNGS (Fig. 3A). Using culture results as a reference, concordance rates were 63.33% (19/30) for wcDNA mNGS and 46.67% (14/30) for cfDNA mNGS. Additionally, the number of pathogenic reads identified by cfDNA mNGS was generally lower than the number identified by wcDNA mNGS across the majority of clinical body fluid specimens (Fig. 3B).

Fig. 3
figure 3

Comparison of cfDNA and wcDNA mNGS relative to culture as a reference. A Cross-tabulation analysis showing the detection agreement between cfDNA mNGS, wcDNA mNGS, and culture. B Comparison of reads of pathogens detected using cfDNA mNGS and wcDNA mNGS. PPV, positive predictive value. NPV, negative predictive value

Full size image

Comparison of pathogen identification by mNGS and 16S rRNA NGS

Forty-one samples were assessed to evaluate the clinical performance of wcDNA mNGS vs. 16S rRNA NGS for detecting bacterial pathogens (Table S1). Results indicated that the sensitivity of wcDNA mNGS were superior to that of 16S rRNA NGS (Fig. 4A). Importantly, wcDNA mNGS showed better consistency with culture results, detecting pathogens in 29 out of 41 samples compared to 24 out of 41 by 16S rRNA NGS. While the correlation between wcDNA mNGS and 16S rRNA NGS results was weak (Pearson correlation coefficient 0.27; p < 0.05), 65% (76/117) of the 16S rRNA NGS results were also detected by wcDNA mNGS (Table S1).

Fig. 4
figure 4

Comparison of bacterial pathogen detection using wcDNA mNGS and 16S rRNA NGS relative to culture as a reference. A Cross-tabulation analysis of wcDNA mNGS and 16S rRNA NGS against culture. B Comparison of reads of pathogens detected by wcDNA mNGS and 16S rRNA NGS. PPV, positive predictive value. NPV, negative predictive value

Full size image

Pathogen identification using wcDNA mNGS in clinical body fluid samples

A total of 125 body fluid samples were analyzed using wcDNA mNGS (Table S1). Among these samples, 54 were aerobic culture-positive, and wcDNA mNGS successfully detected pathogens in 40 of them. Specifically, 27 samples exhibited a single pathogen, while 13 samples contained two or more pathogens. Among the remaining 14 culture-positive samples, four CSF samples and one bile sample yielded monomicrobial pathogen through aerobic culture. The identified pathogens included Acinetobacter baumannii, Staphylococcus haemolyticus, Staphylococcus capitis, Candida tropicalis, and Enterococcus gallinarum. However, the wcDNA mNGS results for these four CSF samples and one bile were as follows: two samples returned negative results, one detected Candida glabrata, another identified Klebsiella pneumoniae, and bile sample D68 detected Enterococcus faecium. Of the nine remaining culture-positive samples, the pathogens isolated by aerobic culture were either not fully detected by wcDNA mNGS, or their read counts were less than 100. For example, bile sample D70 yielded Enterococcus faecium, Serratia marcescens, and Proteus vulgaris using aerobic culture, while wcDNA mNGS detected E. faecium, Clostridium perfringens, and Clostridium baratii. Notably, C. perfringens and C. baratii were also detected by 16S rRNA NGS. In bile sample D83, aerobic culture identified K. pneumoniae and Proteus mirabilis; however, P. mirabilis was not detected by wcDNA mNGS. In sample S24, because wcDNA mNGS detected fewer than 100 reads of Acinetobacter baumannii, the result was negative.

For 71 aerobic culture-negative samples, wcDNA mNGS yielded negative results in 40 cases. However, multiple pathogens were identified in the remaining 31 culture-negative samples. Specifically, K. pneumoniae was detected in five bile samples, while Candida glabrata was identified in eight samples through wcDNA mNGS, including four CSF and ascitic fluid samples. Additionally, two anaerobic bacteria, Parvimonas micra and Prevotella intermedia, were identified using wcDNA mNGS in five samples, which comprised ascites, pleural fluid, and kidney drainage fluid. A. baumannii was identified in four samples from bile, pleural fluid, ascitic fluid, and pancreatic fluid. Moreover, Haemophilus parainfluenzae and other bacteria were detected in four ascites and two CSF samples. Aside from anaerobic bacteria, wcDNA mNGS generally detected low abundances of other pathogens in culture-negative samples (Table S1).

The diagnostic parameters of wcDNA mNGS—sensitivity, specificity, PPV, and NPV—were calculated in comparison to culture, which served as the reference method, as detailed in Table 1. The overall agreement rate between wcDNA mNGS and culture was 64.8%. A chi-square test was performed to assess the association between wcDNA mNGS and culture results, revealing a statistically significant difference (p < 0.05).

Table 1 Clinical performance of wcDNA mNGS in the pathogen detection in body fluid sample
Full size table

Discussion

Culture remains the most commonly used and accepted gold standard for the identification and isolation of pathogenic microbes in clinical microbiology laboratories. However, many studies have confirmed the prevalence of polymicrobial infections, particularly in critically ill or immunocompromised patients [2,3,4]. The culture method is influenced not only by subjective factors, such as the clinician’s experience and technique, but also by objective factors, including the media used, the presence of low-abundance pathogens, and variations in colony morphology [5]. As a result, culture may not effectively identify all potential pathogens in a sample. In contrast, the massively parallel sequencing capabilities of next-generation sequencing allow for the simultaneous detection of multiple pathogens [1, 7,8,9, 13, 15]. In this context, we evaluated the clinical performance of mNGS for pathogen detection and identification in clinical body fluid samples.

Sterile body fluid samples are critical for detecting and diagnosing invasive infectious diseases. These samples are commonly collected from sterile sites, such as CSF, bile, and pancreatic fluid, through procedures like puncture, surgery, or postoperative drainage. Consequently, sterile body fluid samples often contain tissue or cell fragments. Our analysis revealed that whole-cell DNA mNGS outperformed cfDNA mNGS regarding the proportion of host DNA in body fluid samples related to abdominal infections; however, wcDNA mNGS performed poorly in detecting pathogens in CSF samples. Notably, this finding is consistent with a prior study that compared cellular DNA and cfDNA approaches in high host background samples [26]. Four key factors may account for this advantage: First, cfDNA comprises not only fragmented DNA from dying or dead pathogens but also a significant amount of fragmented DNA from degraded host cells. Second, bead beating allows for the comprehensive release of genomic DNA from pathogens contained within host cells or body fluid samples. Third, mNGS provides a full-length genomic analysis, whereas cfDNA is more limiting, focusing on analyzing specific genomic fragments that may not be fully intact. Finally, abdominal infections are often polymicrobial, involving a wide range of microbial species in substantial quantities, leading to a lower proportion of host DNA. When comparing the results of cfDNA mNGS and wcDNA mNGS with clinically relevant culture findings, wcDNA mNGS demonstrated the highest consistency in our cohort. While this finding is consistent with some reports [26,27,28], variability in performance due to factors such as cfDNA extraction methods, sample processing, and bioinformatics analysis underscores the need for further validation across diverse protocols. Furthermore, the detection of pathogens in CSF using cfDNA mNGS remains feasible due to the influence of host DNA, similar to the detection of bloodstream infections using plasma samples [27, 28].

When comparing 16S rRNA NGS to mNGS, notable differences emerged in the identification of pathogenic bacteria. These discrepancies may stem from the semi-quantitative nature of 16S rRNA NGS, influenced by universal degenerate primers and PCR amplification. Nevertheless, mNGS proves capable of identifying a broader range of pathogens and providing a more reliable assessment of pathogen abundance in samples compared to 16S rRNA NGS. Aside from cost considerations, mNGS presents several advantages over 16S rRNA NGS in detecting multiple pathogens.

The clinical performance evaluation of mNGS primarily relies on culture results as a reference, with culture being the leading method in clinical microbiology. However, the inherent limitations of culture should not be overlooked. While high-abundance microorganisms can thrive on specific media, such as blood agar plates, culture may inadvertently overlook those with low abundance or that are difficult to cultivate. To mitigate this challenge, multiple studies have adopted a diversified strategy, integrating culture results with nucleic acid detection and other methods to comprehensively assess mNGS’s accuracy. Despite these efforts, significant challenges remain; not all microorganisms detected in samples are pathogenic, and even combining various techniques cannot fully validate all mNGS results. While mNGS is celebrated for its ability to capture a wide array of microbial sequences—resulting in high sensitivity—it also has relatively low specificity, which can lead to the detection of non-pathogenic microorganisms. In this study, the sensitivity and specificity of mNGS were reported to be 74.07% and 56.34%, respectively. While these results are slightly lower compared to some existing reports on cfDNA mNGS [7, 18], they remain comparable to other studies in the field [13, 22, 29]. Additionally, we found significant differences between the mNGS results and culture results for some samples, particularly as some culture-positive samples were not effectively detected. This could be due to the low abundance of pathogens in these samples or ineffective extraction during the process. Furthermore, culture methods are typically designed for common pathogens, while mNGS is adversely affected by host DNA. This underscores the necessity for careful interpretation of mNGS results in clinical applications and highlights potential research directions aimed at improving mNGS specificity to reduce false positives in the future.

The analysis of body fluid samples reveals several differences in the detection and identification of pathogens between traditional culture methods and wcDNA mNGS. In the case of CSF, pathogen detection via wcDNA mNGS appears to not be more reliable than that of culture [28, 30]. This discrepancy may primarily stem from the influence of host DNA and the typically low abundance of pathogens in CSF, which can hinder accurate detection by mNGS. Additionally, in polymicrobial infections, wcDNA mNGS results demonstrate minor discrepancies with culture results concerning certain low-abundance pathogens, suggesting that while mNGS is a powerful tool, it may not always fully align with culture data [8, 31]. For culture-negative samples, especially from patients with suspected infections, mNGS detection presents a valuable alternative, indicating that wcDNA mNGS might possess broader detection capabilities for certain pathogens, particularly anaerobic organisms that are often challenging to culture [8, 31]. Overall, these findings highlight the complementary roles of culture and mNGS in pathogen detection. Although culture methods remain a cornerstone of microbiological diagnostics, wcDNA mNGS provides significant advantages in identifying a wider array of pathogens, particularly in cases where culture results are negative or limited. Integrating both methods into clinical practice is thus essential for enhancing pathogen identification and improving patient outcomes.

This study has several limitations that should be considered. Firstly, the sample size was modest, indicating the need for larger cohorts in future investigations to enhance the robustness of the findings. Secondly, a critical aspect not addressed in this research was the lack of a comparative analysis between plasma and whole blood samples, as numerous studies have established the utility of plasma microbial cfDNA NGS in detecting and forecasting both systemic and localized infections [13, 14, 16, 17, 19, 20]. To improve the yield of microbial cfDNA and ensure more comprehensive detection, it would be beneficial to utilize larger volumes of supernatant in extraction and concentration processes. Furthermore, while our findings suggest wcDNA mNGS outperformed cfDNA mNGS in this cohort, the use of a single cfDNA extraction kit limits the generalizability of this conclusion. Previous work has demonstrated that variations in cfDNA extraction protocols can significantly impact pathogen detection efficiency [32, 33]. Therefore, future studies employing multiple cfDNA extraction kits and standardized protocols are warranted to validate these observations.

In this study, we collected 125 clinical body fluid samples to evaluate the performance of cfDNA and whole-cell DNA as targets for mNGS, as well as the effectiveness of mNGS and 16S rRNA NGS in identifying pathogens in these samples. Our findings demonstrated that wcDNA mNGS yielded significantly higher sensitivity for pathogen detection and identification compared to both cfDNA mNGS and 16S rRNA NGS in clinical body fluid samples, particularly those associated with abdominal infections. However, the compromised specificity of whole-cell DNA mNGS highlights the necessity for cautious interpretation in clinical settings.

Data availability

The data that support the findings of this study are openly available in National Genomics Data Center at https://ngdc.cncb.ac.cn/gsub/submit/gsa/subCRA016453, reference number [PRJCA016676].

References

  1. Chiu CY, Miller SA. Clinical metagenomics. Nat Rev Genet. 2019;20:341–55.

    CAS  PubMed  PubMed Central  Google Scholar 

  2. Brogden KA, Guthmiller JM, Taylor CE. Human polymicrobial infections. Lancet. 2005;365:253–5.

    PubMed  PubMed Central  Google Scholar 

  3. Tay WH, Chong KKL, Kline KA. Polymicrobial-host interactions during infection. J Mol Biol. 2016;428:3355–71.

    CAS  PubMed  Google Scholar 

  4. Stacy A. The biogeography of polymicrobial infection. Nat Rev Microbiol. 2016;14:93–105.

    CAS  PubMed  Google Scholar 

  5. Sousa AM, Machado I, Nicolau A, Pereira MO. Improvements on colony morphology identification towards bacterial profiling. J Microbiol Methods. 2013;95:327–35.

    PubMed  Google Scholar 

  6. Huang Y, Sheth RU, Zhao S, Cohen LA, Dabaghi K, Moody T, et al. High-throughput microbial culturomics using automation and machine learning. Nat Biotechnol. 2023;41:1424–33.

    CAS  PubMed  PubMed Central  Google Scholar 

  7. Blauwkamp TA, Thair S, Rosen MJ, Blair L, Lindner MS, Vilfan ID, et al. Analytical and clinical validation of a microbial cell-free DNA sequencing test for infectious disease. Nat Microbiol. 2019;4:663–74.

    CAS  PubMed  Google Scholar 

  8. Gu W, Miller S, Chiu CY. Clinical metagenomic next-generation sequencing for pathogen detection. Annu Rev Pathol Mech Dis. 2019;14:319–38.

    CAS  Google Scholar 

  9. Miller S, Naccache SN, Samayoa E, Messacar K, Arevalo S, Federman S, et al. Laboratory validation of a clinical metagenomic sequencing assay for pathogen detection in cerebrospinal fluid. Genome Res. 2019;29:831–42.

    CAS  PubMed  PubMed Central  Google Scholar 

  10. Rodriguez C, Jary A, Hua C, Woerther P-L, Bosc R, Desroches M, et al. Pathogen identification by shotgun metagenomics of patients with necrotizing soft-tissue infections. Br J Dermatol. 2020;183:105–13.

    CAS  PubMed  Google Scholar 

  11. Fida M, Wolf MJ, Hamdi A, Vijayvargiya P, Esquer Garrigos Z, Khalil S, et al. Detection of pathogenic bacteria from septic patients using 16S ribosomal RNA gene-targeted metagenomic sequencing. Clin Infect Dis. 2021;73:1165–72.

    CAS  PubMed  PubMed Central  Google Scholar 

  12. Vos M, Quince C, Pijl AS, de Hollander M, Kowalchuk GA. A comparison of rpoB and 16S rRNA as markers in pyrosequencing studies of bacterial diversity. Gilbert JA, editor. PLoS One. 2012;7:e30600.

    CAS  PubMed  PubMed Central  Google Scholar 

  13. Lee RA, Dhaheri FA, Pollock NR, Sharma TS. Assessment of the clinical utility of plasma metagenomic next-generation sequencing in a pediatric hospital population. J Clin Microbiol. 2020;58:10.

    Google Scholar 

  14. Barrett SLR, Holmes EA, Long DR, Shean RC, Bautista GE, Ravishankar S, et al. Cell free DNA from respiratory pathogens is detectable in the blood plasma of cystic fibrosis patients. Sci Rep. 2020;10:6903.

    CAS  PubMed  PubMed Central  Google Scholar 

  15. Wilson MR, Sample HA, Zorn KC, Arevalo S, Yu G, Neuhaus J, et al. Clinical metagenomic sequencing for diagnosis of meningitis and encephalitis. N Engl J Med. 2019;380:2327–40.

    CAS  PubMed  PubMed Central  Google Scholar 

  16. Farnaes L, Wilke J, Ryan Loker K, Bradley JS, Cannavino CR, Hong DK, et al. Community-acquired pneumonia in children: cell-free plasma sequencing for diagnosis and management. Diagn Microbiol Infect Dis. 2019;94:188–91.

    PubMed  PubMed Central  Google Scholar 

  17. Goggin KP, Gonzalez-Pena V, Inaba Y, Allison KJ, Hong DK, Ahmed AA, et al. Evaluation of plasma microbial cell-free DNA sequencing to predict bloodstream infection in pediatric patients with relapsed or refractory cancer. JAMA Oncol. 2020;6:552.

    PubMed  Google Scholar 

  18. Gu W, Deng X, Lee M, Sucu YD, Arevalo S, Stryke D, et al. Rapid pathogen detection by metagenomic next-generation sequencing of infected body fluids. Nat Med. 2021;27:115–24.

    CAS  PubMed  Google Scholar 

  19. Wang L, Guo W, Shen H, Guo J, Wen D, Yu Y, et al. Plasma microbial cell-free DNA sequencing technology for the diagnosis of sepsis in the ICU. Front Mol Biosci. 2021;8: 659390.

    CAS  PubMed  PubMed Central  Google Scholar 

  20. Xu C, Chen X, Zhu G, Yi H, Chen S, Yu Y, et al. Utility of plasma cell-free DNA next-generation sequencing for diagnosis of infectious diseases in patients with hematological disorders. J Infect. 2023;86:14–23.

    CAS  PubMed  Google Scholar 

  21. Sun N, Chen Y, Zhang J, Cao J, Huang H, Wang J, et al. Identification and characterization of pancreatic infections in severe and critical acute pancreatitis patients using 16S rRNA gene next-generation sequencing. Front Microbiol. 2023;14: 1185216.

    PubMed  PubMed Central  Google Scholar 

  22. Yu J, Zhang L, Gao D, Wang J, Li Y, Sun N. Comparison of metagenomic next-generation sequencing and blood culture for diagnosis of bloodstream infections. Front Cell Infect Microbiol. 2024;14: 1338861.

    CAS  PubMed  PubMed Central  Google Scholar 

  23. Lu J, Rincon N, Wood DE, Breitwieser FP, Pockrandt C, Langmead B, et al. Metagenome analysis using the Kraken software suite. Nat Protoc. 2022;17:2815–39.

    CAS  PubMed  PubMed Central  Google Scholar 

  24. Yan G, Liu J, Chen W, Chen Y, Cheng Y, Tao J, et al. Metagenomic next-generation sequencing of bloodstream microbial cell-free nucleic acid in children with suspected sepsis in pediatric intensive care unit. Front Cell Infect Microbiol. 2021;11: 665226.

    CAS  PubMed  PubMed Central  Google Scholar 

  25. Breitwieser FP, Salzberg SL. Pavian: interactive analysis of metagenomics data for microbiome studies and pathogen identification. Schwartz R, editor. Bioinformatics. 2020;36:1303–4.

    CAS  PubMed  Google Scholar 

  26. Chen H, Zheng Y, Zhang X, Liu S, Yin Y, Guo Y, et al. Clinical evaluation of cell-free and cellular metagenomic next-generation sequencing of infected body fluids. J Adv Res. 2024;55:119–29.

    CAS  PubMed  Google Scholar 

  27. Lu Y, Zhang Y, Lou Z, He X, Zhang Q, Zhang Q, et al. Metagenomic next-generation sequencing of cell-free DNA for the identification of viruses causing central nervous system infections. Microbiol Spectr. 2024;12: e0226423.

    PubMed  Google Scholar 

  28. Li X, Yang L, Li D, Yang X, Wang Z, Chen M, et al. Diagnosis of neurological infections in pediatric patients from cell-free DNA specimens by using metagenomic next-generation sequencing. Microbiol Spectr. 2023;11: e0253022.

    PubMed  Google Scholar 

  29. Serpa PH, Deng X, Abdelghany M, Crawford E, Malcolm K, Caldera S, et al. Metagenomic prediction of antimicrobial resistance in critically ill patients with lower respiratory tract infections. Genome Med. 2022;14:74.

    CAS  PubMed  PubMed Central  Google Scholar 

  30. Yu L, Zhang Y, Zhou J, Zhang Y, Qi X, Bai K, et al. Metagenomic next-generation sequencing of cell-free and whole-cell DNA in diagnosing central nervous system infections. Front Cell Infect Microbiol. 2022;12: 951703.

    CAS  PubMed  PubMed Central  Google Scholar 

  31. Miao Q, Ma Y, Wang Q, Pan J, Zhang Y, Jin W, et al. Microbiological diagnostic performance of metagenomic next-generation sequencing when applied to clinical practice. Clin Infect Dis. 2018;67:S231–40.

    CAS  PubMed  Google Scholar 

  32. Terp SK, Pedersen IS, Stoico MP. Extraction of cell-free DNA: evaluation of efficiency, quantity, and quality. J Mol Diagn. 2024;26:310–9.

    CAS  PubMed  Google Scholar 

  33. Oreskovic A, Brault ND, Panpradist N, Lai JJ, Lutz BR. Analytical comparison of methods for extraction of short cell-free DNA from urine. J Mol Diagn. 2019;21:1067–78.

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

This study was funded by the Suqian Sci&Tech Program (No. SY202214), the National Natural Science Foundation of China (No. 81601857), and the Health Technology Development Special Foundation of Nanjing City (No. YKK18216).

Author information

Author notes

  1. Ning Sun and Jiaxun Zhang contributed equally to this work.

Authors and Affiliations

  1. Department of Clinical Laboratory, Jinling Hospital, Affiliated Hospital of Medical School, Nanjing University, Nanjing, China

    Ning Sun, Jin Cao, Yong Chen, Deyu Gao & Xinyi Xia

  2. Department of Clinical Laboratory, Anhui Provincial Chest Hospital (Provincial Institute of Tuberculosis Control), Hefei, Anhui, China

    Jiaxun Zhang

  3. Department of Microbiology and Immunology, College of Basic Medicine, Guangdong Medical University, Dongguan, China

    Wentao Guo

  4. State Key Laboratory of Analytical Chemistry for Life Science, Nanjing University, Nanjing, Jiangsu, 210093, China

    Xinyi Xia

Authors
  1. Ning Sun
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Jiaxun Zhang
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Wentao Guo
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Jin Cao
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Yong Chen
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Deyu Gao
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Xinyi Xia
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

Conceptualization: Ning Sun, Xinyi Xia; Formal analysis: Ning Sun, Jiaxun Zhang, Xinyi Xia; Funding acquisition: Ning Sun, Xinyi Xia; Resources: Ning Sun, Jiaxun Zhang, Wentao Guo, Jin Cao, Yong Chen; Experimental studies: Jiaxun Zhang, Jin Cao, Yong Chen, Deyu Gao; Supervision: Sun Ning, Xinyi Xia, Wentao Guo; Writing-original draft: Jiaxun Zhang, Ning Sun, Xinyi Xia; Revising-original draft: Ning Sun, Xinyi Xia. All authors read and accepted the final version of the manuscript.

Corresponding authors

Correspondence to Ning Sun or Xinyi Xia.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the Human Use Ethical Committee at Jinling Hospital (approval number: 2021DZGZR-YBB-013), and the residual samples were collected for retrospective analysis after routine testing in our clinical laboratory. All experimental procedures were by the ethical standards of Jinling Hospital of China and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all participants or, where applicable, from their legal guardians.

Consent for publication

Not applicable.

Competing of interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sun, N., Zhang, J., Guo, W. et al. Comparative analysis of metagenomic next-generation sequencing for pathogenic identification in clinical body fluid samples. BMC Microbiol 25, 165 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12866-025-03887-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12866-025-03887-8

Keywords

BMC Microbiology

ISSN: 1471-2180

Contact us