Skip to main content
Download PDF

Integrating DNA and RNA sequencing for enhanced pathogen detection in respiratory infections

Journal of Translational Medicine volume 23, Article number: 325 (2025) Cite this article

This article has been updated

Abstract

Background

The clinical value of shotgun metagenomic next-generation sequencing (mNGS) in improving the detection rates of respiratory pathogens is well-established. However, mNGS is complex and expensive. This study designed and evaluated the performance of targeted NGS (tNGS) in diagnosing respiratory infections.

Methods

We retrospectively included samples from 281 patients with lower respiratory tract infections to establish thresholds of pathogens. Subsequently, target pathogens were selected and a probe hybridization system was established. The performance and clinical manifestations of tNGS for 306 pathogens were evaluated using clinical and simulated samples.

Results

The tNGS method took 16 h with sequencing data sizes of 5 M reads. The limit-of-detection of tNGS was 100–200 CFU/mL, respectively. Bioinformatics simulation confirmed the method’s high specificity and robustness. In 281 patients of clinical validation cohort, tNGS exhibited a sensitivity of 97.73% and specificity of 75.41% compared to the composite reference standard, which notably surpasses those of culture-based and conventional microbiological methods (CMT). In detecting bacterial and viral infection, tNGS demonstrated superior sensitivity relative to CMT. Notably, 61.40% of target viruses were subtype-resolved with the initial establishment of reliable typing cutoffs, with the subtyping results being completely consistent with the PCR results. tNGS allowed for concurrent identification of antimicrobial resistance (AMR) markers and viral subtyping. 80.56% of AMR markers identified by tNGS were consistent with antimicrobial susceptibility testing.

Conclusion

This research established the robust performance of our tailored tNGS assay in the simultaneous detection of DNA and RNA pathogens, underscoring its prospective suitability for widespread use in clinical diagnostics.

Introduction

Lower respiratory tract infections (LRTIs) ranked as the 4th leading cause of morbidity and mortality among both adults and children, resulting in a death toll of 2.6 million in 2019 [1]. Early pathogen identification was the key to targeting effective therapy and avoiding abuse of broad-spectrum antibiotics [2,3,4]. Due to some pathogens being difficult or impossible to in vitro culture, the etiologies of up to 62% of community-acquired pneumonia remain undiagnosed despite comprehensive diagnostic work-up [5,6,7].

Approaches based on polymerase chain reaction (PCR) that identify pathogens directly from the specimen can speed identification but these tests include only a small number of pathogens, requiring a presumptive diagnosis before a test is chosen [8, 9]. Metagenomic next-generation sequencing (mNGS) may serve as a new tool to overcome the shortcomings of conventional diagnostic methods [10]. The advantage of mNGS was that it could simultaneously identify all potential etiology in a sample without presetting. Several studies have explored the advantages of mNGS in pathogen detection for LRTIs patients, and found that mNGS resulted in an increase of 15–30% in pathogen detection and a shorter time to diagnosis [11,12,13]. Separate DNA or RNA detection could inherently predispose this method to the risk of missing RNA pathogens [14,15,16]. Meanwhile, respiratory tract samples usually contain large amounts of human nucleic acid [17], leading to insufficient coverage for the detection of true pathogens and limiting the comprehensive analysis of drug-resistance genes and subtype analysis [18].

Targeted NGS (tNGS) utilized pathogen-specific primers or probes to enrich for targeted microorganisms, enabling efficient and cost-effective detection, while maintaining sensitivity and specificity comparable to metagenomic next-generation sequencing (mNGS) for lower respiratory tract pathogens [14,15,16]. Precise design of the tNGS target panel facilitated the identification and analysis of antimicrobial resistance (AMR) markers and pathogen variants [19]. Capture probe enrichment, an approach employing hybridization with capture bait probes to targeted metagenomic libraries [20], was adapted in this study to develop and validate a tNGS assay capable of simultaneous DNA and RNA detection. The rationale for simultaneous detection of DNA and RNA pathogens is primarily due to the atypical clinical manifestations or lack of conventional detection methods for some RNA viruses, such as human metapneumovirus [21]. Additionally, infections by atypical DNA pathogens are difficult to distinguish in the early stages based on symptoms alone, which may lead to delayed treatment. This is exemplified by pathogens such as Mycoplasma pneumoniae [22], Legionella spp [21]., and Pneumocystis jirovecii [23, 24]. Recent studies have confirmed that conventional clinical detection methods are insufficient for identifying these pathogens. However, broad-spectrum NGS methods, such as mNGS and tNGS, have demonstrated significant potential as a valuable supplement [25,26,27]. Therefore, this present study aims to screen for both potential DNA and RNA pathogens in a single test while concurrently assessing AMR genes.

Methods

Study design and sample enrollment

All samples were collected between July 2022 and July 2023 from the Laboratory of Clinical Microbiology and Infectious Diseases of China-Japan Friendship Hospital, and Shanghai Key Laboratory of Lung Inflammation and Injury of Zhongshan Hospital, Fudan University. Samples were obtained from suspected LRTI patients after provider-ordered testing. Residual samples were required to meet minimal volume requirements for enrollment, including at least one swab sample, 2 ml of BALF, or 2 ml of sputum. To evaluate the performance of the tNGS assay in detecting RNA viruses, we included swab samples from suspected SARS-CoV-2 infected patients.

Conventional Microbiological tests

Various conventional microbiological tests (CMTs) methods were applied as part of provider-ordered testing, including Gram stain and bacterial culture, fungal stain and culture, galactomannan antigen testing, mycobacterial stain and culture, Mycobacterium tuberculosis by the GeneXpert MTB/RIF system (GeneXpert; Cepheid, Inc., Sunnyvale, CA, USA), respiratory virus multiplex PCR, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) PCR. The above tests were selected in the diagnosis according to patient symptoms. Demographic and clinical patient data were collected through the hospital’s electronic medical record system. Adult patients with suspected LRTI, who have complete clinical information and are non-duplicate, were included (Fig. 1).

Fig. 1
figure 1

Flowchart of experimental design and tNGS workflow. (A) Workflow of clinical analysis for pathogen diagnosis. Samples were assessed using both tNGS and CMT, with the results compared against the gold standard (composite reference standard). Discrepant results were confirmed using mNGS or quantitative PCR (qPCR). Finally, threshold establishment for tNGS and analysis of its clinical performance were conducted. (B) All nucleic acid was extracted including bacteria, fungi, viruses, and human; RNA was reverse transcribed into cDNA and all DNA was used for library preparation. Then, libraries were captured and enriched by DNA baits; Finally, sequencing and analysis were performed. Summary diagram of the experimental process time for tNGS assays(bottom)

Full size image

Targeted NGS panel design and evaluation of probe hybridization capture system

This tNGS technology is an improved method based on the construction of conventional mNGS libraries. Its probe hybridization capture process can specifically bind to microorganisms, enhancing the detection of microorganisms. In this study, an enrichment panel including 306 pathogens or subtypes related to respiratory tract infections was designed (Supplementary Table S1). The selection of these pathogens referred to medical books, expert consensus on respiratory tract infections, articles, databases, and opinions of frontline experts [28, 29](Supplementary Fig. S1). The reference database was primarily curated from the NCBI RefSeq/nt, where redundant sequences with high similarity and low-quality data were removed to construct a pathogen reference genome database. Based on the reference genomes, we designed a comprehensive capture probe panel, which includes species-specific genes, conserved species genes, and associated antimicrobial resistance genes of the pathogens.

A systematic evaluation of the probe hybridization conditions was conducted and validated the optimal probe dosage among 0.2, 0.3, 0.4, and 0.5 fmol using simulated samples. It was found that 0.3 fmol exhibited the best performance in this reaction system(Supplementary Fig. S2). Additionally, we explored the optimal temperatures for hybridization, capture, and washing, which were determined to be 60 °C, 65 °C, and 70 °C, respectively. Although longer hybridization capture times yielded better performance, the clinical requirement for a shorter turnaround time (TAT) in pathogen detection led us to apply a 4-hour hybridization in this project, which is comparable to an 8-hour hybridization and slightly better than a 1-hour hybridization.

Sample processing and Dual-nucleic acid extraction

Sputum from 100 µl was mixed with 3 times of liquefied reagent (Vazyme, Nanjing, China) for 10–15 min at 37 ℃, and then centrifuged at 12,000×g for 5 min. The dried oropharyngeal swab samples were soaked in 1 ml of sterile PBS for 10 min. BALF samples with high viscosity were diluted 1:1 with 0.1 M dithiothreitol before nucleic acid extraction. Then, the 400µL of samples (BALF, sputum, and swab), lysis buffer, protease K mixture, binding buffer, and 1.2 g glass bead were agitated vigorously at 4500 rpm for a total of 130 s by FastPrep-24™ 5G Instrument (MP Biomedical, CA, USA). The separated 0.25 mL sample was used to extract nucleic acids using the VAMNE Magnetic Pathogen DNA/RNA kit (Vazyme, Nanjing, China). Then, the RNA was reverse-transcribed into cDNA using Hieff NGS® ds-cDNA Synthesis Kit (Yeasen, Shanghai, China). Nucleic acids were quantified using an Invitrogen™ Qubit™ 3.0 fluorometer(Thermo Fisher, Waltham, MA, USA) to ensure all samples with concentration ≥ 0.1 ng/µL or else subjected for library reconstruction. A549 human cells (GenePlus, Suzhou, China) were used as negative controls (NTC) to detect contamination, and A549 human cells spiked with Staphylococcus aureus (BeNa Culture Collection, Beijing, China) were used as positive controls (PTC).

Library Preparation and enrichment

Load 50 ng of the extracted nucleic acid based on its concentration. The HieffNGS®C37P4 One Pot cDNA&gDNA Library Prep Kit (Yeasen, Shanghai, China) was used for cDNA synthesis and library preparation following the provided protocol. Then, the cDNA was taken through library enrichment with NadPrep® NanoBlockers (Nanodigmbio, Nanjing, China) reagents to generate the product for targeted sequencing. Eight uniquely barcoded libraries were pooled to hybridize and capture by specific biotinylated probes for 4 h using self-design probes. In the probe hybridization capture process, 0.3 fmol of probes were added, and hybridization and capture were performed at 60 °C and 65 °C for 4 h. Subsequently, elution was conducted at 70 °C to complete the hybridization capture and enrichment process. Products were quantified with an Invitrogen™ Qubit™ 3.0 fluorometer (Thermo Fisher, Waltham, MA, USA) to ensure all samples with concentration ≥ 1 ng/µL. Products were stored at − 20 °C until Sequencing.

Sequencing and bioinformatic analysis

Sequencing was performed on Gene + Seq-100 (GenePlus, Suzhou, China) with a 100-bp single-end read sequencing to goal depths of 5 million reads per sample. After sequencing, ensure that the final sequencing output is no less than 5 million reads, with Q20 and Q30 values not lower than 95% and 88%, respectively. Clean reads were obtained by removing sequencing adapters, low-quality reads, or reads below 35 bp using fastp (version 0.23.1) [30]. The remaining reads were aligned to the human reference (hg38) using Burrow-Wheeler Aligner (version 0.7.12-r1039) [31], and human reads were filtered. The filtered reads were compared with the self-built pathogenic microorganism database, and the retained results were annotated. SARS-CoV-2 subtype classification was accomplished using both Nextclade and Pangolin tools [32] (Fig. 1B). The number of reads per one million sequencing reads (RPM) was calculated at the species and genus levels. The thresholds were established based on the results of this study. The thresholds for viruses, bacteria, and fungi were 9.90, 5.44, and 0.667 RPM, respectively.

Analytical performance

The detection limits (LoD) of eight representative microorganisms (Fig. 2A) were assessed. These eight microorganisms were obtained from the National Institutes for Food and Drug Control. The microorganisms were spiked into A549 human cells(105 cells/mL) at descending concentrations of 10,000, 1,000, 500, 200, 100, and 50 CFU/mL (for viruses, copies/mL) respectively. If a concentration of 50 CFU/mL was detectable, microorganism dilution continued until it was undetectable. Three replicates were performed for each concentration. The LoD is defined as the RPM values of the microorganisms exceeding a predetermined threshold.

Fig. 2
figure 2

Analytical of limit-of-detection and interference testing. (A) The limit-of-detection of different spiked organism. Eight representative microorganisms were spiked into concentrations of 10,000, 1,000, 500, 200, 100, and 50 CFU/mL (for viruses, copies/mL) respectively. (B) Confirmation test of LoD by a 23 microorganisms panel. Each microorganisms were tested three times at 1-fold LoD. The left side represents the results for all samples, while the right side shows the results for all pathogens. (C) Human source sequence interference testing. Three concentration gradients of 104, 105, and 106 cells/mL, respectively, with a concentration of 1000 copies/mL (CFU/mL) for the four pathogens were used. (D) Interference testing for closely related species detection. The reported DNA concentrations are shown for two pairs of similar organisms, S. epidermidis and S. aureus, S. pneumoniae and S. mitis when coinfection was contrived at ratios of 1:99, 1:9,1:1, 9:1 and 99:1. The expected DNA concentration was calculated from samples where each organism was spiked into a sample alone. Staphylococcus epidermidis was not covered by tNGS

Full size image

A reference panel composed of 23 microorganisms (listed in Supplementary Table S2) selected from the CHINET (www.chinets.com), WHO list of priority fungi and China CDC report(www.chinacdc.cn/) was designed to assess the limit-of-detection (LoD) for the tNGS assay. A purified solution of 23 microorganisms was obtained from BeNa Culture Collection.

Precision evaluation samples were contrived in the 1-fold, 3-fold, and 10-fold LoD. Three aliquots were thawed and processed according to the standard workflow on each day of testing, and three runs were tested in 3 days. The within-run reproducibility detected three replicates of the same sample processed in parallel during the same run.

A set of representative microorganisms (Staphylococcus aureus/Klebsiella pneumoniae/respiratory syncytial virus/Candida albicans) was mixed with different A549 human cell densities (104, 105, 106 cells/mL) to analyze host interference.

To analyze the coinfection interference among microorganisms with high genomic similarity within the same genus, mixtures with different mixing ratios of closely related species within the same genus (S. aureus/S. epidermidis, Streptococcus pneumoniae/S. mitis) were used. Five groups of mixtures were prepared.

Bioinformatic cross-reactivity

To validate the specificity and completeness of the microorganisms database, robustness tests at varying concentrations of different microorganisms within the database were conducted. Fragments of each microorganism at concentrations near LoD and 10-fold LoD were extracted from the predefined microorganisms’ target regions, and a sample was constructed for each microorganism. To simulate mixed infections and assess database specificity, mixed genomic samples of 3 microorganisms at 10-fold LoD and 10 microorganisms at 10-fold LoD were extracted, and the specificity of their annotations was simulated. All these samples were mixed with the host to 5 M data.

Diagnostic assessment

The extra orthogonal testing was performed on samples with inconsistent results between tNGS and CMT results, such as qPCR experiments for virus (Liferiver, Shanghai, China), GeneXpert MTB/RIF assays (Xpert-MTB), and mNGS(GenePlus, Suzhou, China). Two clinically experienced physicians assessed patient infection status based on clinical presentation, imaging, CMT, and tNGS results. Subsequently, combined with laboratory and clinical data, pathogen results from the Karius test are categorized as definite, probable, possible, or unlikely based on the comprehensive reference criteria outlined [33]. In various case scenarios, the definite and probable pathogens are further classified into pathogenic categories based on their clinical significance. The clinical sensitivity and specificity of targeted Next-Generation Sequencing (tNGS) are calculated and compared with conventional microbiological testing (CMT) and clinical adjudication.

Statistics

Inter-group comparisons are made using the unpaired t-test or the Mann-Whitney U test. For comparisons between groups of categorical variables, the chi-square test is used. Statistical comparisons were conducted using SPSS version 26.0 (SPSS Inc., Chicago, Illinois, USA). All figures were drawn using GraphPad Prism (version 9.2.0) and R 4.3.1 software (http://www.r-project.org, The R Foundation). Error calculations represent 95% confidence intervals or standard deviations, where indicated. Statistical significance was set at a p-value of < 0.05.

Results

Analytical performance for tNGS

The detection timeline for tNGS in this study is depicted in Fig. 1B. The entire tNGS process takes approximately 16 h, which is comparable to mainstream mNGS methods [34].

Eight representative microorganisms were selected for evaluation, as the limit-of-detection (LoD) cannot be established for all analytes covered by the tNGS panel [14]. The performance was depicted at different concentrations of representative organism pools (Fig. 2A). The LoD of bacteria, viruses, and fungi were 100 CFU/mL, 100 copies/mL, and 200 CFU/mL, respectively. The LoD for the microorganisms was validated using a reference panel composed of 23 microorganisms. The composition of these 23 microorganisms was similar to that of the 306 panel, and they had comparable GC content (Supplementary Fig. S3). Both the 23 pathogens and NTCs were tested in triplicate at one-fold LoD, and all pathogens and samples were correctly identified, demonstrating a sensitivity and specificity of 100% (Fig. 2B).

The samples contained different concentrations of representative organisms to measure the precision and coefficient of variation (CV). The tNGS assay demonstrated high within- and between-run reproducibility (100%). The average coefficient of variation (CV) for the within-run was 20.34%, and for the between-run, it was 26.42% (Table 1).

Table 1 Within run and Between-run precision of tNGS tests
Full size table

The interference was assessed using samples with varying host loads but the same representative microbial load(Fig. 2C). As the host load increased, there was some fluctuation in the RPM values of the microbes. The results of the tNGS detection were slightly affected by the host load.

Analytical specificity of tNGS

One of the highest risk factors for false positives in bioinformatics is cross-reactivity between genetically similar microorganisms when the DNA content of one microbe is high. This cross-reactivity can occur when genomic regions can map to multiple organisms or reference genomes in the database. We first tested the risk of alignment false positives through simulated samples at different levels (Table 2). Initially, the performance of high-level infection(10-fold LoD) samples from 306 microorganisms was assessed. All 306 microorganisms could be identified, with a positive predictive value (PPV) of 99.67% (an additional false positive was detected in one sample, belonging to the same genus as the true positive call). Considering the diversity of clinical microbial genomes, we introduced certain variations in the simulated samples to mimic the differences between reference genomes and actual clinical microorganisms. At high levels of infection (10-fold LoD), consistent with previous results, the PPV was 99.67%. Only one false positive appeared in the near LoD simulated sample, with a PPV of 99.67%, and the overall specificity of the analysis was 99.999%. In consideration of the potential for mixed infections in actual samples to lead to false positives in calling, we further simulated samples with multiple microorganism mixtures. In samples with 3 mixed microorganisms, all 300 microorganisms were identified, with a positive predictive value (PPV) of 99.67% (an additional false positive from the same genus was identified in one sample). In samples with 10 mixed microorganisms, all 1000 microorganisms were still identifiable, but 2 samples identified additional false positives, resulting in a PPV of 99.80%.

Table 2 The diversity robustness of bioinformatics database
Full size table

The final aspect of analytical specificity investigated was interference during coinfection with genetically similar organisms. Two pairs, S. epidermidis/ S. aureus and S. pneumoniae/ S. mitis, were tested in various mix ratios. Results showed that despite co-infection, the microbial DNA concentration remained comparable to mono-infection expectations (Fig. 2D). However, tNGS did not detect S. epidermidis due to its coverage in the panel.

Diagnostic performance of tNGS

A retrospective cohort ultimately included 281 samples from 281 patients, comprising 192 bronchoalveolar lavage fluid (BALF) samples, 69 sputum samples, and 20 swab samples. 50.18% of these patients originated from the intensive care unit. The baseline information of the patients is presented in Supplementary Table S3.

In this cohort, we established thresholds using 261 BALF and sputum samples and evaluated these thresholds with 192 BALF samples. The thresholds for viruses, bacteria, and fungi were set at 9.90, 5.44, and 0.667 Reads Per Million (RPM), respectively. The tNGS method demonstrated good diagnostic performance, with the area under the receiver operating characteristic curve (ROC) for all pathogens exceeding 0.83.

As shown in Fig. 3A and 42.70% of samples were positive in culture, 67.97% of samples were positive by CMT, and 82.21% of samples were detected for microorganisms by tNGS. Compared with culture (Fig. 3), tNGS assay had a sensitivity of 96.67%(95%CI 91.18-98.93%), and a specificity of 33.33%(95%CI 25.76-41.83%). Additional microbiological testing including cultures of other specimens, serology, and nucleic acid testing, identified 191 positives. 186 was identified by tNGS demonstrating a sensitivity of 97.38%(95%CI 93.66-99.03%) and a specificity of 52.22%(95%CI 41.49-62.76%). In comparison to the composite reference standard, tNGS sequencing demonstrated a sensitivity of 97.73%(95%CI 94.49-99.16%), and a specificity of 75.41%(95%CI 60.69-83.83%). Overall, the identification of infections alert etiology was significantly higher for tNGS test (215 of 281) than for culture (120 of 261, p < 0.001) and CMT combined (190 of 281, p < 0.001)( Supplementary Fig. S4).

Fig. 3
figure 3

Clinical validity. (A) Receiver operating characteristic (ROC) curves showing the diagnostic value of tNGS in BALF and sputum. (B) The proportion of samples identified as causal pathogens by different methods. (C) Contingency tables for tNGS results vs. clinical culture, conventional microbiological testing, and composite reference standard. NA: not available. PPV: positive predictive value

Full size image

Then, we assessed the performance of the tNGS assay across bacterial, fungal, and viral infections, demonstrating sensitivities exceeding 95% and superiority over CMT for bacterial and viral infection(p < 0.05), but similar in fungal infection (Fig. 3B, Supplementary Fig. S4). High sensitivity was also found in three specimens (Fig. 3C). Samples from swabs, sputum, and BALF all exhibited sensitivities above 95%.

Fig. 4
figure 4

Detection performance of pathogens using tNGS. (A) Venn plot showing the overlap of causal pathogens detected among CMT and tNGS. (B) Comparison of normalized reads in the causal pathogen group (TP) and non-causal pathogen group (FP). (C) Comparison of causal pathogen detection between CMT and tNGS. The bar chart illustrates the capability of tNGS and CMT in detecting each pathogen. The sunburst chart summarizes the total number of detections for each taxon(bacteria, fungi, and virus) by both methods. (D). Heat map showing consistency among the tNGS methods, clinical diagnosis and CMT

Full size image

In the five false negative samples for tNGS, 80% (4/5) were also false negative of mNGS(Supplementary Table S4). In addition, among the 56 patients with mixed infections, tNGS and CMT provided diagnostic results for 55 and 53 patients respectively. Among the 61 samples for which no definite or probable microbiological cause of the infection alert was identified, no significant microbial DNA was detected in 46 by tNGS sequencing. Organisms DNA was identified by tNGS in 15 samples likely to be commensals which were also identified by other molecular methods (Supplementary Table S5).

Diagnostic performance of tNGS in pathogens

Further analysis of the diagnostic performance of tNGS for various pathogens revealed that in BALF samples, the sensitivity of tNGS exceeded 90% for Gram-negative bacteria (GNB), Gram-positive bacteria (GPB), fungi, viruses, and atypical pathogens (Table 3). For example, the sensitivity to Streptococcus pneumoniae reaches 100%; The sensitivity to Klebsiella pneumoniae, P. jirovecii, Human Metapneumovirus and Mycoplasma pneumoniae both reach 100%. However, in sputum samples, the sensitivity for Staphylococcus aureus and Acinetobacter baumannii was below 90%, at 81.82% and 87.50%, respectively. A total of 285 pathogens were identified in this cohort (Fig. 4A). tNGS detected a significantly higher number of pathogens compared to CMT (p = 0.001), particularly in the detection of GNB and viruses (Fig. 4B). It is important to note, however, that tNGS may also identify some false positives, which show no significant difference in RPM from true positive pathogens (Fig. 4C). tNGS has demonstrated superior detection capabilities for atypical pathogens, identifying a greater number of Aspergillus fumigatus, P. jirovecii, M. abscessus, Legionella pneumophila, and Chlamydophila psittaci (Fig. 4C&D).

Table 3 Diagnostic performance of tNGS in pathogens
Full size table

Performance of tNGS on RNA-virus subtype recognition

Conscious probe design enables the method to perform subtyping analysis of RNA viruses. This study specifically collected throat swab samples for analysis. A total of 57 samples from the overall cohort were identified as SARS-CoV-2, with 61.40%(35/57) of SARS-CoV-2 samples further subjected to subtyping analysis and consistent with the results of qPCR. Analysis indicated that an RPM > 119 allows for the analysis of variant strains, and ROC value was 0.9934 (Supplementary Fig. S5).

Evaluation of tNGS performance in predicting antimicrobial susceptibility

The tNGS panel and bioinformatic analysis also covered and enabled the detection of bacterial genotypic AMR markers associated with antimicrobial susceptibility testing (AST) results for certain agents. Thirty-six samples had the results of standard-of-care methods detected which phenotypic resistance had been evaluated and the AMR marker was compared with these results (Supplementary Table S6). Agreement between the associated and tested resistance was found in 29 of 36 (80.56%) potential pathogens, including carbapenem-resistant A. baumannii, methicillin-resistant S. aureus (MRSA), and M. tuberculosis with first-line agents’ resistance. All four first-line drug-resistant M.tuberculosis s were found to harbor associated AMR genes, and 5 of 7 MRSA isolates tested positive for the mecA gene. In two-thirds of the K. pneumoniae and 9 of 11 of the A. baumannii resistance genes consistent with phenotypic resistance were identified.

Discussion

Lower respiratory tract infections could be caused by a multitude of pathogens, with viral pneumonias, particularly after SARS-CoV-2 amidst the global pandemic, garnering significant attention. While mNGS assays contribute to the identification of infectious pathogens, their separate handling of DNA and RNA may lead to the underdetection of RNA viruses. In this study, we designed and validated a tNGS assay capable of concurrent DNA and RNA detection. We assessed the performance of this tNGS assay using simulated laboratory samples (comprising wet-lab experiments and bioinformatic simulations) and retrospective clinical specimens. The tNGS had demonstrated satisfactory clinical performance.

Organism types were detected by the NGS with differing efficiencies [14]. The detection of viruses and bacteria was higher in mNGS [35, 36], and that of fungi was varied [34, 37]. We attempted to evaluate the LoD and precision of all organisms through a combination of traditional and metagenomic-specific validation strategies. Like other reports [14, 38], the viruses and bacteria were diminished by the LoD of 100 CFU/mL to detect organisms at lower abundances, and 200 CFU/mL in fungi. In a recent study, the tNGS showed an LoD ranging between 103 and 104 CFU/mL similar to mNGS [14]. A lower LOD was not entirely harmless, and in our analysis, we found that the PPV of tNGS assay was lower than CMT in bacteria and viruses. As in previous studies [24], this research also observed that a high concentration of host nucleic acids can, to some extent, affect the capture efficiency of probes, leading to fluctuations in the RPM values of target microorganisms. This is reflected as a relatively high CV of the microorganisms’ RPM values in the precision assessment, both within and between run. The same results have been observed in other studies, with the CV values of both tNGS and mNGS exceeding 20% in different detections [24, 33]. Despite this impact, both this study and others have confirmed that it has minimal impact on the correct identification of microorganisms, even those near the detection limit. However, this also reminds us that in clinical practice, we should pay attention to microorganisms near the threshold, as fluctuations in RPM may lead to “missed detections”.

In the retrospective analysis, tNGS demonstrated higher agreement with culture and CMT methods and identified a greater number of aetiological causes of respiratory tract infections than standard-of-care testing [11, 38]. This study confirms that tNGS outperforms culture and CMT in etiological diagnosis, aligning with prior mNGS findings [34, 36]. tNGS exhibited superior diagnostic performance for bacterial and viral infections, with comparable fungal detection to CMT. Molecular detection of fastidious fungi like Aspergillus spp. enhances tracking and prevention of clinical fungal infections [12]. The tNGS method has demonstrated superior sensitivity in detecting several clinically significant pathogens. Specifically, tNGS has shown higher sensitivity (100%) for the detection of M. tuberculosis and non-tuberculous mycobacteria, as well as for L. pneumophila, M. pneumoniae, and C. psittaci. Culturing is not an ideal method for the identification of these pathogens. The utilization of serological testing and PCR methods is influenced by the selection of the attending physician. Broad-spectrum NGS, such as mNGS, is regarded as an appropriate method for screening and detection of these pathogens [22, 23]. The detection rates for these pathogens using tNGS are higher than those achieved by the molecular detection methods included in the CMT approach. This highlights the potential of tNGS in detecting fastidious and atypical pathogens, which can be particularly beneficial in clinical settings where accurate and rapid diagnosis is crucial. However, the sensitivity and breadth of microorganisms detected, combined with the diversity of the microbiome compositions in respiratory tract specimens, makes it challenging to achieve high diagnostic specificity [11]. Some false positive samples were validated by mNGS in which the microbial DNA also was found in most samples. Therefore, the potential pathogens identified in 24.59% of negative samples was the microecology of the human, due to its conditional pathogenicity, it must be reported under the tNGS rules. Some false negatives also were validated by mNGS and the interference of commensal microorganisms in the respiratory tract is one of the reasons for this result. Rapid detection of AMR was also essential in critically ill patients with severe bacterial infections given that time to appropriate antimicrobials correlates with mortality [39, 40]. In our cohort, 80.56% of agreement was found in 36 samples between AMR and AST which may play an auxiliary role in the diagnosis and treatment choice of clinical patients.

The study has confirmed the potential of tNGS for viral genotyping based on probe hybridization capture methods. As SARS-CoV-2 continues to ravage globally, the correlation between viral subtypes and patient damage and prognosis is being increasingly recognized by medical professionals [41]. Meanwhile, members of our team participated in a comparison of the diagnostic values of tNGS and mNGS in patients with severe pneumonia and immunocompromised individuals [25, 27]. The results showed that tNGS and mNGS demonstrated similar clinical diagnostic performance. Moreover, the turnaround time of tNGS was one - third shorter than that of mNGS (median 16 h vs. 24 h). Additionally, the cost of tNGS was significantly reduced, being only half that of mNGS ($250 vs. $500) [24]. tNGS, as a broad-spectrum NGS method that is more cost-effective (only half the price of mNGS) [24], may in the future be used for the evolutionary surveillance of epidemic pathogens. Future research could consider integrating tNGS with artificial intelligence (AI). Studies have confirmed that combining NGS sequencing results with machine learning can significantly enhance the accuracy of predicting antimicrobial resistance phenotypes [42, 43]. Moreover, the integration of AI with NGS technology can also be utilized to predict viral evolutionary trends [44]. The combination of NGS and AI may play a more significant role in the future, particularly in the rapid diagnosis of severe infections, personalized antibiotic guidance, and public health infection control [45].

This study has multiple limitations. First, in the initial assessment of tNGS’s clinical performance, it is important to note that the inclusion of samples did not strictly adhere to the principles of randomization. Second, orthogonal testing, such as qPCR, was not performed on all samples in parallel. Third, the number of organisms tested was limited, resulting in an inadequate evaluation of certain types of pathogens. Fourth, the number of microorganisms covered by the tNGS panel designed in this study is limited. It mainly takes into account the common pathogens of respiratory tract infections. In the future, we are considering developing a tNGS method that covers a greater number of pathogenic microorganisms based on this research, which will be applicable to a wider range of infectious syndromes.

Conclusions

In summary, the detection capability of tNGS for pathogens in respiratory samples has been systematically evaluated, demonstrating that this method can meet the needs of most clinical scenarios for infection diagnosis.

Data availability

The data that support the findings of this study are available from the corresponding author on request. Sequencing data that support the finding of this study (with human reads removed) have been deposited in the China National Center for Bioinformation - National Genomics Data Center and can be accessed with the BioProject identifier PRJCA022061.

Change history

  • 12 April 2025

    A typesetting mistake in affiliations 4 and 6 has been corrected.

Abbreviations

tNGS:

Targeted Next-generation sequencing

RTIs:

Respiratory tract infections

AMR:

Antimicrobial resistance

AST:

Antimicrobial susceptibility testing

mNGS:

Metagenomic next-generation sequencing

BALF:

Bronchoalveolar lavage fluid

ROC:

Receiver operating characteristic

LoD:

Limit-of-detection

PPV:

Positive predictive value

NPV:

Negative predictive value

CMT:

Conventional microbiological test

References

  1. GLC. Age-sex differences in the global burden of lower respiratory infections and risk factors, 1990–2019: results from the global burden of disease study 2019. Lancet Infect Dis. 2022;22(11):1626–47.

    Article  Google Scholar 

  2. Barlam TF, Cosgrove SE, Abbo LM, MacDougall C, Schuetz AN, Septimus EJ, et al. Implementing an antibiotic stewardship program: guidelines by the infectious diseases society of America and the society for healthcare epidemiology of America. Clin Infect Diseases: Official Publication Infect Dis Soc Am. 2016;62(10):e51–77.

    Article  Google Scholar 

  3. Liesenfeld O, Lehman L, Hunfeld KP, Kost G. Molecular diagnosis of sepsis: new aspects and recent developments. Eur J Microbiol Immunol. 2014;4(1):1–25.

    Article  CAS  Google Scholar 

  4. Kumar A, Ellis P, Arabi Y, Roberts D, Light B, Parrillo JE, et al. Initiation of inappropriate antimicrobial therapy results in a fivefold reduction of survival in human septic shock. Chest. 2009;136(5):1237–48.

    Article  PubMed  Google Scholar 

  5. Jain S, Self WH, Wunderink RG, Fakhran S, Balk R, Bramley AM, et al. Community-Acquired pneumonia requiring hospitalization among U.S. Adults. N Engl J Med. 2015;373(5):415–27.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  6. Nelson JC, Jackson M, Yu O, Whitney CG, Bounds L, Bittner R, et al. Impact of the introduction of Pneumococcal conjugate vaccine on rates of community acquired pneumonia in children and adults. Vaccine. 2008;26(38):4947–54.

    Article  PubMed  Google Scholar 

  7. Griffin MR, Zhu Y, Moore MR, Whitney CG, Grijalva CG. U.S. Hospitalizations for pneumonia after a decade of Pneumococcal vaccination. N Engl J Med. 2013;369(2):155–63.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Buchan BW, Windham S, Balada-Llasat JM, Leber A, Harrington A, Relich R et al. Practical Comparison of the BioFire FilmArray Pneumonia Panel to Routine Diagnostic Methods and Potential Impact on Antimicrobial Stewardship in Adult Hospitalized Patients with Lower Respiratory Tract Infections. Journal of clinical microbiology. 2020;58(7).

  9. Zhu Y, Xu B, Li C, Chen Z, Cao L, Fu Z, et al. A multicenter study of viral aetiology of Community-Acquired pneumonia in hospitalized children in Chinese Mainland. Virol Sin. 2021;36(6):1543–53.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  10. Chiu CY, Miller SA. Clinical metagenomics. Nat Rev Genet. 2019;20(6):341–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  11. 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 Diseases: Official Publication Infect Dis Soc Am. 2018;67(suppl2):S231–40.

    Article  CAS  Google Scholar 

  12. Liang M, Fan Y, Zhang D, Yang L, Wang X, Wang S, et al. Metagenomic next-generation sequencing for accurate diagnosis and management of lower respiratory tract infections. Int J Infect Diseases: IJID: Official Publication Int Soc Infect Dis. 2022;122:921–9.

    CAS  Google Scholar 

  13. Qu J, Zhang J, Chen Y, Huang Y, Xie Y, Zhou M, et al. Aetiology of severe community acquired pneumonia in adults identified by combined detection methods: a multi-centre prospective study in China. Emerg Microbes Infections. 2022;11(1):556–66.

    Article  CAS  Google Scholar 

  14. Gaston DC, Miller HB, Fissel JA, Jacobs E, Gough E, Wu J, et al. Evaluation of metagenomic and targeted Next-Generation sequencing workflows for detection of respiratory pathogens from Bronchoalveolar lavage fluid specimens. J Clin Microbiol. 2022;60(7):e0052622.

    Article  PubMed  Google Scholar 

  15. Li S, Tong J, Liu Y, Shen W, Hu P. Targeted next generation sequencing is comparable with metagenomic next generation sequencing in adults with pneumonia for pathogenic microorganism detection. J Infect. 2022;85(5):e127–9.

    Article  CAS  PubMed  Google Scholar 

  16. Deng X, Achari A, Federman S, Yu G, Somasekar S, Bártolo I, et al. Metagenomic sequencing with spiked primer enrichment for viral diagnostics and genomic surveillance. Nat Microbiol. 2020;5(3):443–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Diao Z, Han D, Zhang R, Li J. Metagenomics next-generation sequencing tests take the stage in the diagnosis of lower respiratory tract infections. J Adv Res. 2022;38:201–12.

    Article  CAS  PubMed  Google Scholar 

  18. Li N, Cai Q, Miao Q, Song Z, Fang Y, Hu B. High-Throughput metagenomics for identification of pathogens in the clinical settings. Small Methods. 2021;5(1):2000792.

    Article  CAS  PubMed  Google Scholar 

  19. Xiao M, Liu X, Ji J, Li M, Li J, Yang L, et al. Multiple approaches for massively parallel sequencing of SARS-CoV-2 genomes directly from clinical samples. Genome Med. 2020;12(1):57.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. García-García G, Baux D, Faugère V, Moclyn M, Koenig M, Claustres M, et al. Assessment of the latest NGS enrichment capture methods in clinical context. Sci Rep. 2016;6:20948.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Wei Y, Zhang T, Ma Y, Yan J, Zhan J, Zheng J, et al. Clinical evaluation of metagenomic Next-Generation sequencing for the detection of pathogens in BALF in severe community acquired pneumonia. Ital J Pediatr. 2023;49(1):25.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Liang Y, Dong T, Li M, Zhang P, Wei X, Chen H, et al. Clinical diagnosis and etiology of patients with Chlamydia psittaci pneumonia based on metagenomic next-generation sequencing. Front Cell Infect Microbiol. 2022;12:1006117.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Li X, Li Z, Ye J, Ye W. Diagnostic performance of metagenomic next-generation sequencing for Pneumocystis jirovecii pneumonia. BMC Infect Dis. 2023;23(1):455.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Yin Y, Zhu P, Guo Y, Li Y, Chen H, Liu J, et al. Enhancing lower respiratory tract infection diagnosis: implementation and clinical assessment of multiplex PCR-based and hybrid capture-based targeted next-generation sequencing. EBioMedicine. 2024;107:105307.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Zhang P, Liu B, Zhang S, Chang X, Zhang L, Gu D, et al. Clinical application of targeted next-generation sequencing in severe pneumonia: a retrospective review. Crit Care (London England). 2024;28(1):225.

    Article  Google Scholar 

  26. Chen Q, Yi J, Liu Y, Yang C, Sun Y, Du J et al. Clinical diagnostic value of targeted next–generation sequencing for infectious diseases (Review). Mol Med Rep. 2024;30(3).

  27. Wei M, Mao S, Li S, Gu K, Gu D, Bai S, et al. Comparing the diagnostic value of targeted with metagenomic next-generation sequencing in immunocompromised patients with lower respiratory tract infection. Ann Clin Microbiol Antimicrob. 2024;23(1):88.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Li ZJ, Zhang HY, Ren LL, Lu QB, Ren X, Zhang CH, et al. Etiological and epidemiological features of acute respiratory infections in China. Nat Commun. 2021;12(1):5026.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Liu YN, Zhang YF, Xu Q, Qiu Y, Lu QB, Wang T, et al. Infection and co-infection patterns of community-acquired pneumonia in patients of different ages in China from 2009 to 2020: a National surveillance study. Lancet Microbe. 2023;4(5):e330–9.

    Article  PubMed  Google Scholar 

  30. Chen S, Zhou Y, Chen Y, Gu J. Fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinf (Oxford England). 2018;34(17):i884–90.

    Google Scholar 

  31. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinf (Oxford England). 2009;25(14):1754–60.

    CAS  Google Scholar 

  32. Aksamentov I, Roemer C, Hodcroft E, Neher R. Nextclade: clade assignment, mutation calling and quality control for viral genomes. J Open Source Softw. 2021;6:3773.

    Article  Google Scholar 

  33. 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(4):663–74.

    Article  CAS  PubMed  Google Scholar 

  34. Peng JM, Du B, Qin HY, Wang Q, Shi Y. Metagenomic next-generation sequencing for the diagnosis of suspected pneumonia in immunocompromised patients. J Infect. 2021;82(4):22–7.

    Article  CAS  PubMed  Google Scholar 

  35. Liu H, Zhang Y, Yang J, Liu Y, Chen J. Application of mNGS in the etiological analysis of lower respiratory tract infections and the prediction of drug resistance. Microbiol Spectr. 2022;10(1):e0250221.

    Article  PubMed  Google Scholar 

  36. Sun T, Wu X, Cai Y, Zhai T, Huang L, Zhang Y, et al. Metagenomic Next-Generation sequencing for pathogenic diagnosis and antibiotic management of severe Community-Acquired pneumonia in immunocompromised adults. Front Cell Infect Microbiol. 2021;11:661589.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  37. Lin P, Chen Y, Su S, Nan W, Zhou L, Zhou Y, et al. Diagnostic value of metagenomic next-generation sequencing of Bronchoalveolar lavage fluid for the diagnosis of suspected pneumonia in immunocompromised patients. BMC Infect Dis. 2022;22(1):416.

    Article  PubMed  PubMed Central  Google Scholar 

  38. Murphy CN, Fowler R, Balada-Llasat JM, Carroll A, Stone H, Akerele O et al. Multicenter evaluation of the biofire filmarray pneumonia/pneumonia plus panel for detection and quantification of agents of lower respiratory tract infection. J Clin Microbiol. 2020;58(7).

  39. 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(1):74.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Lee CC, Lee CH, Yang CY, Hsieh CC, Tang HJ, Ko WC. Beneficial effects of early empirical administration of appropriate antimicrobials on survival and defervescence in adults with community-onset bacteremia. Crit Care (London England). 2019;23(1):363.

    Article  Google Scholar 

  41. Zhang J, Dong P, Liu B, Xu X, Su Y, Chen P, et al. Comparison of XBB and BA.5.2: differences in clinical characteristics and disease outcomes. Arch Bronconeumol. 2023;59(11):782–4.

    Article  PubMed  Google Scholar 

  42. Zhou X, Yang M, Chen F, Wang L, Han P, Jiang Z, et al. Prediction of antimicrobial resistance in Klebsiella pneumoniae using genomic and metagenomic next-generation sequencing data. J Antimicrob Chemother. 2024;79(10):2509–17.

    Article  CAS  PubMed  Google Scholar 

  43. Tian Y, Zhang D, Chen F, Rao G, Zhang Y. Machine learning-based colistin resistance marker screening and phenotype prediction in Escherichia coli from whole genome sequencing data. J Infect. 2024;88(2):191–3.

    Article  CAS  PubMed  Google Scholar 

  44. Nie Z, Liu X, Chen J, Wang Z, Liu Y, Si H, et al. A unified evolution-driven deep learning framework for virus variation driver prediction. Nat Mach Intell. 2025;7(1):131–44.

    Article  Google Scholar 

  45. Wong F, de la Fuente-Nunez C, Collins JJ. Leveraging artificial intelligence in the fight against infectious diseases. Sci (New York NY). 2023;381(6654):164–70.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

We gratefully acknowledge Professor Yuanlin Song of Shanghai Institute of Infectious Disease and Biosecurity for his guidance and advice on the study. We owe thanks to the patients in our study and their family members. We acknowledge the staffs of all centers for their assistance to this study.

Funding

This work was supported by Beijing Natural Science Foundation (L222073), Capital’s funds for health improvement and research (CFH, 2024-1-4063), Shanghai Municipal Science and Technology Major Project (ZD2021CY001), and Science and Technology Commission of Shanghai Municipality (20DZ2261200).

Author information

Author notes

  1. Dejian Gu, Jie Liu, Jiaping Wang, Jun She and Binghuai Lu contributed equally to this work.

Authors and Affiliations

  1. Geneplus-Beijing Co., Ltd., Beijing, China

    Dejian Gu, Rui Gao & Hao Liu

  2. Shanghai Key Laboratory of Lung Inflammation and Injury, Department of Pulmonary and Critical Care Medicine, Zhongshan Hospital, Fudan University, Shanghai, China

    Jie Liu & Jun She

  3. Suzhou GenePlus Clinical Laboratory Co., Ltd, Beijing, China

    Jiaping Wang, Yuting Yi & Yuxing Chu

  4. Laboratory of Clinical Microbiology and Infectious Diseases, Department of Pulmonary and Critical Care Medicine, Center of Respiratory Medicine, Beijing Key Laboratory of Surveillance, Early Warning and Pathogen Research on Emerging Infectious Diseases, National Center for Respiratory Medicine, China-Japan Friendship Hospital, Beijing, China

    Binghuai Lu

  5. Shanghai Key Laboratory of Lung Inflammation and Injury, Department of Pulmonary and Critical Care Medicine, Zhongshan Hospital, Fudan University, 180 Fenglin Road, Shanghai, 200032, China

    Jun She

  6. Laboratory of Clinical Microbiology and Infectious Diseases, Department of Pulmonary and Critical Care Medicine, China-Japan Friendship Hospital, No. 2 East Yinghua Street, Chaoyang District, Beijing, 100029, China

    Binghuai Lu

Authors
  1. Dejian Gu
    View author publications

    You can also search for this author inPubMed Google Scholar

  2. Jie Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  3. Jiaping Wang
    View author publications

    You can also search for this author inPubMed Google Scholar

  4. Yuting Yi
    View author publications

    You can also search for this author inPubMed Google Scholar

  5. Yuxing Chu
    View author publications

    You can also search for this author inPubMed Google Scholar

  6. Rui Gao
    View author publications

    You can also search for this author inPubMed Google Scholar

  7. Hao Liu
    View author publications

    You can also search for this author inPubMed Google Scholar

  8. Jun She
    View author publications

    You can also search for this author inPubMed Google Scholar

  9. Binghuai Lu
    View author publications

    You can also search for this author inPubMed Google Scholar

Contributions

BL, JS, and YY contributed to conception and design of the study. JW, JL, and YC contributed to the operation of the experiments and the collection of information. DG, HL, and RG contributed to confirmation of the authenticity of the data. DG, JL and JW performed data analysis and interpretation. All authors wrote the manuscript and read and approved the final manuscript.

Corresponding authors

Correspondence to Jun She or Binghuai Lu.

Ethics declarations

Ethics approval and consent to participate

This study was approved by the institutional review boards at the China-Japan Friendship Hospital with approval 2023-KY-099.

Consent for publication

Not applicable.

Competing interests

Dejian Gu, Yuting Yi, Yuxing, Chu, Jiaping, Wang, Rui, Gao, and Hao Liu are employees of Beijing GenePlus Co., Ltd., and have participated in the development and experimental processes of the entire workflow. The other authors declare that they have no competing interests.

Additional information

Publisher’s note

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

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

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

Gu, D., Liu, J., Wang, J. et al. Integrating DNA and RNA sequencing for enhanced pathogen detection in respiratory infections. J Transl Med 23, 325 (2025). https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-025-06342-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doiorg.publicaciones.saludcastillayleon.es/10.1186/s12967-025-06342-4

Keywords

Journal of Translational Medicine

ISSN: 1479-5876

Contact us