A fully integrated sample-to-answer diagnostic platform enables rapid quantification of post-transplant infection at the point-of-care Authors: Pengkun Yin, Lu Gao, Hao Zhang, Jun Zeng, Dan Huang, Lifan Yang, Saifu Yin, Turun Song, Feng Li Published: 2026 ABSTRACT Opportunistic infections after organ transplantation require frequent monitoring over years. However, current diagnostic tests are inaccessible to patients in resource-limited settings. Here, we present a microfluidic-enhanced nucleic acid test for on-site reporting (MONITOR), a fully integrated, sample-in-answer-out platform for point-of-care (POC) detection of opportunistic infections. MONITOR employs a gene-specific DNAzyme that, upon integration with recombinase polymerase amplification, enables CRISPR-like trans-cleavage signal amplification with single-molecule sensitivity. It further incorporates a rapid, extraction-free protocol for processing the urine sample, lyophilized assay reagents for cold-chain-free storage and transportation, and a 3D-printed microfluidic cartridge for self-contained, contamination-free fluid handling. A portable real-time fluorescence reader and smartphone application are further developed to enable on-site quantification. Clinical validation showed complete concordance with qPCR across the tested cohorts and a field trial identified high-risk patients with systemic BK polyomavirus infection. MONITOR fulfills ideal criteria for POC testing, offering a practical solution for decentralized, quantitative viral monitoring in low-resource settings. Introduction Kidney transplantation remains the treatment of choice for end-stage renal diseases, however, the 10-year graft survival rates stagnate less than 50%1–3. Opportunistic infections, particularly the BK polyomavirus (BKV), represent a major cause of allograft loss, necessitating life-long monitoring of post-transplant patients4–7. While conventional urinalysis for decoy cell offers limited clinical specificity8, the direct quantification of viral load via quantitative polymerase chain reaction (qPCR) has emerged into the gold standard for diagnosing opportunistic infections9–11. Nevertheless, current qPCR-based diagnostics relies on expensive and bulky equipment, multistep workflows, and well-trained personnel, limiting their accessibility, particularly for kidney transplant recipients (KTRs) living in underserved areas12. Moreover, due to the long-distance travel, slow laboratory processing and turnaround time in clinical settings, it typically takes several days to weeks to obtain the test results, leading to diagnostic delays and increased risks of irreversible allograft injury13,14. Therefore, there is a critical need for affordable, point-of-care (POC) tests that are accessible to KTRs in remote or resource-limited settings. Such tools would enable more frequent and rapid monitoring, facilitating earlier detection and intervention for infections and graft dysfunc An ideal diagnostic test for the frequent monitoring of opportunistic infections in POC settings shall combine the sensitivity and specificity of nucleic acid testing with the speed and simplicity of antigen-based tests, such as lateral flow tests15–17. CRISPR Cas systems offer a promising path toward such stringent requirements and have already enabled several field-deployable nucleic acid tests for viral outbreaks, including SARS-CoV-218–22, Zika23–25, and dengue virus26–28. For instance, Collins and colleagues demonstrated rapid detection of BKV and cytomegalovirus (CMV) DNA from blood and urine samples of KTRs by integrating CRISPR Cas13 with recombinase polymerase amplification (RPA)29. Despite many CRISPR-based and other portable nucleic acid tests have been widely validated and employed for clinical tests in real-world field conditions, those for field-based detection of opportunistic infections in KTRs remain underexplored30–37. To bridge this gap, key innovations are needed to match the practical demands of on-site testing in resource-limited settings, including 1) stable reagent formulations that permit long-term storage and cold-chain-free transportation; 2) field-compatible sample processing protocol that ideally bypasses nucleic acid extraction; 3) full integration of all test steps, from sample processing to signal reporting, into a single device with no external liquid transfer to improve ease of use and prevent carryover contamination; and 4) quantitative readout capabilities to support accurate viral load assessment and infection risk stratification. Herein, we report a microfluidic-enhanced nucleic acid test for on-site reporting (MONITOR), designed for the POC detection of opportunistic infections post-transplantation under field conditions (Fig. 1). We first develop a CRISPR-like assay based on screening and chemically modifying gene-specific DNAzymes (gDzs) that enables signal amplification upon specific activation by BKV DNA. This protein-free trans-cleavage system offers not only comparable sensitivity to that of CRISPR Cas machinery but also remarkably lower cost and greater stability at room temperature, enabling a shelf-life for more than 1 year and cold-chain-free transportation over distances of more than 4000 km. We further develop an improved heating unextracted diagnostic samples to obliterate nuclease (HUDSON) protocol, allowing the direct release of viral DNA from unprocessed urine samples within 3 min. The entire MONITOR workflow, including in situ urine sample processing, nucleic acid amplification via RPA, and signal reporting mediated by gDz, was integrated into a microfluidic device equipped with self-contained cartridges and a rotary valve as the sequencing control unit, enabling a single-step, sample-to-answer test within 1 h. A compact fluorescence reader with multiple temperature zones and smartphone-based application (APP) are also developed to facilitate real-time fluorescence monitoring and quantitative analysis. MONITOR achieves 100% sensitivity (95% CI, 94.9–100%) and specificity (95% CI, 92.7–100%) in a clinical validation against 120 clinical samples collected from KTRs, which included 69 samples processed with standard nucleic acid extraction, 36 unprocessed urine samples and 15 unprocessed blood samples. Finally, to validate MONITOR in a real-world resource-limited setting, we conduct a field trial 400 kilometers from the central laboratory at West China Hospital of Sichuan University. Urine samples from 21 KTRs are analyzed on-site with MONITOR, promptly identifying 3 high-risk KTRs requiring close follow up and further validation. Of note, subsequent urine and blood testing in central laboratory confirmed one of these patients had a systemic BKV infection necessitating clinical intervention.Fig. 1MONITOR vs. conventional workflow for monitoring opportunistic infections in kidney transplant recipients.Because of the long travel distance and slow turnaround time in central laboratories, it typically takes more than 1 week for KTRs to receive the diagnostic report. By integrating all testing steps into a single microfluidic cartridge, MONITOR allows on-site detection of opportunistic infections from unprocessed urine samples within 1 h. MONITOR vs. conventional workflow for monitoring opportunistic infections in kidney transplant recipients. Because of the long travel distance and slow turnaround time in central laboratories, it typically takes more than 1 week for KTRs to receive the diagnostic report. By integrating all testing steps into a single microfluidic cartridge, MONITOR allows on-site detection of opportunistic infections from unprocessed urine samples within 1 h. Results Design of MONITOR assay for BKV detection We began by screening gDzs for BKV using a selection protocol previously established by our lab38. Specifically, a classic 10–23 RNA-cleaving Dz was reproposed for BKV analysis by keeping its 15-nt catalytic core but screening binding arm sequences against the genomic DNA of BKV (Fig. 2a and Supplementary Fig. 1). The arm lengths were designed to be 6–8 nt on each site to ensure both high catalytic efficiency and sequence specificity to the target gene. We designed and tested 4 gDzs targeting the coding region of the small T antigen (STA) which possessed sequence homology of more than 95% among all BKV strains accessible from the National Center for Biotechnology Information (NCBI). Results in Fig. 2b and Supplementary Fig. 2 revealed BKV-Dz-2 as the optimal gDz with an apparent rate constant (kobs) of 1.003 min−1.Fig. 2Design of gene-specific DNAzyme (gDz)-based MONITOR assay for BKV detection.a Screening of four BKV-specific gDzs derived from the 10–23 DNAzyme by varying the binding-arm sequences according to the BKV genome. b Evaluating the catalytic activities of BKV-specific gDz. The bands correspond to the intact 16-nt substrate (S) and the 9-nt cleavage product (P). The observed rate constants kobs were extracted from the gel image (n = 1). c Schematic illustration of engineering a BKV-specific gDz into a target-responsive sensor for amplified BKV DNA detection. d Evaluating the sensitivity of BKV sensor by plotting fluorescence signals of cleaved reporters at 60 min against target DNA at varying concentrations. P value was calculated by two-tailed Student’s t-test. e Screening of chemical modifications to enhance gDz-sensor activity. Modification 4 contained 2′-O-methyl substitutions at positions 7 and 8, a 2′-methoxyethyl substitution at position 14 and phosphorothioate linkages at positions 1 and 13. Data are displayed as mean ± SD (n = 3). f Workflow of the MONITOR assay for the ultrasensitive detection of BKV DNA via RPA, post-amplification assembly, and gDz-based recognition and trans-cleavage of fluorogenic reporters. g Kinetic curves for analyzing synthetic BKV DNA standards at concentrations from 6 × 102 to 6 × 107 cp/mL in original samples. Data are displayed as mean ± SD (n = 3). h Evaluating the sensitivity of MONITOR assay by plotting endpoint fluorescence (30 min) against target concentrations. Data are displayed as mean ± SD (n = 3). P values were calculated by two-tailed Student’s t-test. i Evaluating the long-term stability of the MONITOR assay solution, including the gDz sensor, the reporter, and helper strands at ambient temperature for 30 days. Data are displayed as mean ± SD (n = 3). Source data are provided as a Source Data file. Design of gene-specific DNAzyme (gDz)-based MONITOR assay for BKV detection. a Screening of four BKV-specific gDzs derived from the 10–23 DNAzyme by varying the binding-arm sequences according to the BKV genome. b Evaluating the catalytic activities of BKV-specific gDz. The bands correspond to the intact 16-nt substrate (S) and the 9-nt cleavage product (P). The observed rate constants kobs were extracted from the gel image (n = 1). c Schematic illustration of engineering a BKV-specific gDz into a target-responsive sensor for amplified BKV DNA detection. d Evaluating the sensitivity of BKV sensor by plotting fluorescence signals of cleaved reporters at 60 min against target DNA at varying concentrations. P value was calculated by two-tailed Student’s t-test. e Screening of chemical modifications to enhance gDz-sensor activity. Modification 4 contained 2′-O-methyl substitutions at positions 7 and 8, a 2′-methoxyethyl substitution at position 14 and phosphorothioate linkages at positions 1 and 13. Data are displayed as mean ± SD (n = 3). f Workflow of the MONITOR assay for the ultrasensitive detection of BKV DNA via RPA, post-amplification assembly, and gDz-based recognition and trans-cleavage of fluorogenic reporters. g Kinetic curves for analyzing synthetic BKV DNA standards at concentrations from 6 × 102 to 6 × 107 cp/mL in original samples. Data are displayed as mean ± SD (n = 3). h Evaluating the sensitivity of MONITOR assay by plotting endpoint fluorescence (30 min) against target concentrations. Data are displayed as mean ± SD (n = 3). P values were calculated by two-tailed Student’s t-test. i Evaluating the long-term stability of the MONITOR assay solution, including the gDz sensor, the reporter, and helper strands at ambient temperature for 30 days. Data are displayed as mean ± SD (n = 3). Source data are provided as a Source Data file. BKV-Dz-2 was then engineered into a BKV-specific sensor by further optimizing its arm lengths and then blocking its arm sequences using a complementary motif (Fig. 2c and Supplementary Fig. 3). The complementary motif contained identical sequence as the substrate except that the rGrU ribonucleotides were replaced by the deoxyribonucleotide counterparts to prevent cleavage. Melting analysis and structure prediction proved that the folded gDz-sensor was highly stable at room temperature or 37 °C, ensuring no signal leakage in the absence of the BKV DNA (Supplementary Fig. 4). A toehold-domain was also designed so that the blocked gDz could be rapidly activated by BKV DNA. Upon activation, gDz unleashed its catalytic activity for amplified cleavage of fluorophore-quencher dual labeled substrate sequences with a limit of detection (LOD) estimated at 0.1 nM (Fig. 2c, d, Supplementary Fig. 5 and Supplementary Fig. 6). Kinetic analysis further revealed remarkably higher catalytic activity of our gDz over a classic multicomponent Dz design (Supplementary Fig. 7). As several previous studies suggested chemical modifications at certain positions could help enhance catalytic activities of 10–23 Dzs39–42, we also examined the effects of 7 types of chemical modifications on our gDz sensor (Supplementary Fig. 8). Results in Fig. 2e and Supplementary Fig. 9 revealed that the combination (modification-4) of 2‘-O-methyl modification at positions 7 and 8, 2‘-methoxyethyl (MOE) modification at position 14, and 2 phosphorothioate linkages at positions 1 and 13 was most effective to enhance the target-induced trans-cleavage activity of BKV-Dz-2 without increasing the target-independent background signal. We next integrated gDz sensor with isothermal RPA to achieve the sensitivity necessary for BKV detection in real clinical samples. RPA primers were adopted from previous designs by Collins and colleagues29, which were compatible with the chemically modified BKV-gDz-2 for the specific detection of BKV at the STA gene (Supplementary Fig. 10). A post-amplification assembly protocol that converted double-stranded RPA amplicons into DNA assemblies with single-stranded detection windows using a panel of short helper strands via snap cooling was also introduced, enabling the rapid analysis of RPA amplicons using the gDz sensor (Fig. 2f and Supplementary Fig. 11). By combining RPA and gDz, the MONITOR assay allowed the detection of synthetic BKV DNA standard down to 6 × 102 cp/mL (Fig. 2g and h), corresponding to 1.5 molecules in the original sample (2.5 µL). This assay sensitivity represents single-molecular detection level comparable to CRISPR Cas13-based SHERLOCK assay. Moreover, unlike CRISPR assay relying on fragile protein components, our MONITOR assay is a protein-free detection system with remarkably enhanced reagent stability. Even after storing at room temperature for 30 days, gDz sensor remained highly efficient for target-induced signal amplification with minimal signal leakage (Fig. 2i). Improving the accessibility of MONITOR assay For achieving the frequent monitoring of opportunistic infection post-transplantation under decentralized conditions, improvements were needed to increase user-friendliness and facilitate test distribution. Towards this goal, we incorporated ultrafast and extraction-free urine sample processing and lyophilized test reagents into our MONITOR assay, thereby decreasing the overall complexity of the assay and enabling cold-chain-free transportation and long-term storage. To achieve extraction-free sample processing for MONITOR, three requirements must be met, including 1) viral particles must be effectively lysed to release BKV DNA, 2) nucleases must be inactivated as the reporter contained two ribonucleotides, and 3) the protocol must be simple and fast enough to be compatible with the microfluidic system. HUDSON has been demonstrated to be a highly effective nuclease inactivation and viral lysis method for CRISPR-based assays but requires incubation at two temperatures and 20–30 min incubation28. To be compatible with MONITOR, we sought to modify the HUDSON protocol to achieve fast and single-step processing of urine samples without compromising the assay performance (Fig. 3a). The concentration of Tris (2-carboxyethyl) phosphine (TCEP) was found to be a key factor that influenced both processing time and assay performance. For urine sample analysis, high concentrations of TCEP (100 mM) used in the original HUDSON protocol were found to significantly reduce the sensitivity of the MONITOR assay, whereas 100 mM remained to be optimal for blood sample analysis (Supplementary Fig. 12). By reducing the TCEP concentration to 10 mM, we were able to achieve comparable detection signals for BKV DNA spiked in urine samples to those in buffer systems (Fig. 3b, and Supplementary Fig. 13). No significant impact was found for the MONITOR assay from common urinary interferents, such as bilirubin, hemoglobin, and glucose, at varying concentrations or combinations (Supplementary Fig. 14 and Supplementary Fig. 15). Moreover, the reduced TCEP concentration allowed effective viral release using a single incubation step at 95 °C. More excitingly, this modified HUDSON protocol for MONITOR was ultrafast, allowing extraction-free sample processing and sufficient nuclease inactivation within only 3 min (Fig. 3c and d).Fig. 3Improving the accessibility of the MONITOR assay via extraction-free sample processing and reagent lyophilization.a Schematic illustration of the one-step, extraction-free processing of urine sample to release BKV DNA. b Evaluation of the MONITOR assay performance for urine samples undergone varying treatment protocols. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. c Heatmap by plotting fluorescence signals of the MONITOR assay against varying target concentrations and heating time during sample processing. d Evaluating the nuclease activity using RNase Alert kit before or after treating the urine sample with HUDSON or MONITOR protocol. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. e Long-term stability of lyophilized gDz sensor mixed with reporters at ambient temperature for over 60 weeks. For test at each time point, the target concentration was fixed at 10 nM. f Evaluating the MONITOR assay performance after long-distance distribution of lyophilized reagents to 7 cities without the use of cold chain transportation. Source data are provided as a Source Data file. Improving the accessibility of the MONITOR assay via extraction-free sample processing and reagent lyophilization. a Schematic illustration of the one-step, extraction-free processing of urine sample to release BKV DNA. b Evaluation of the MONITOR assay performance for urine samples undergone varying treatment protocols. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. c Heatmap by plotting fluorescence signals of the MONITOR assay against varying target concentrations and heating time during sample processing. d Evaluating the nuclease activity using RNase Alert kit before or after treating the urine sample with HUDSON or MONITOR protocol. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. e Long-term stability of lyophilized gDz sensor mixed with reporters at ambient temperature for over 60 weeks. For test at each time point, the target concentration was fixed at 10 nM. f Evaluating the MONITOR assay performance after long-distance distribution of lyophilized reagents to 7 cities without the use of cold chain transportation. Source data are provided as a Source Data file. Distribution and storage of test reagents under resource-limited conditions are often barriers that limit POC diagnostics in real clinical settings. Lyophilization can facilitate the long-term storage and cold-chain-free distribution of diagnostic reagents to remote areas. As the commercial RPA reagent was provided as the lyophilization powder already, we optimized lyophilization protocol for the gDz sensor. We found that the addition of 1% trehalose to the lyophilization buffer could effectively enhance stability of the gDz sensor and improve solubility upon rehydration. The removal of magnesium ion (Mg2+) was also found to be critical to ensure the stability of the ribonucleotide-containing reporter probes. Under optimal conditions, both gDz and the reporter probe could be lyophilized together with full activity and no observable signal leakage at room temperature for over 1 year (Fig. 3e, Supplementary Fig. 16 and Fig. 17). The much-improved shelf life of lyophilized MONITOR reagent over those of lyophilized CRISPR diagnostic reagents (e.g., SHINEv.2 with a shelf life only over 1 week20) echoed the advantage of our protein-free assay design. We further challenged lyophilized MONITOR reagents for long-distance distribution under real-world settings. To do so, three test tubes containing lyophilized gDz sensor, RPA reagents, and helper strands, respectively, were sealed together into a vacuum-sealed aluminum-plastic packaging bag and sent on a round trip from Chengdu to 7 destinations of varying distances under ambient temperatures. Results in Fig. 3f and Supplementary Fig. 18 revealed stable performance for analyzing 6 × 104 cp/mL synthetic BKV DNA in all MONITOR tests even after transporting for up to 4,000 kilometers and 7 days without cold chain. Collectively, the combination of on-site sample processing, long-term storage at ambient temperature, and cold-chain-free transportation ensures the accessibility of our MONITOR assay to KTRs located at remote and/or resource-constrained areas. Enhancing user-friendliness of MONITOR via a self-contained and sequence-controllable microfluidic cartridge To enable frequent monitoring of opportunistic infections in POC settings, the MONITOR assay must be easily operated by regular healthcare personnel with no specialized training. To achieve this goal, we integrated all operational steps into a self-contained and sequence-controllable microfluidic cartridge (Fig. 4a and b). The cartridge contained four independent storage chambers, allowing the pre-load of all sample processing and nucleic acid testing reagents necessary for the MONITOR assay. A rotary valve was designed to achieve the sequential addition of reagents into a reaction chamber, where connection between each storage chamber and the reaction chamber was established by controlling the angle of the rotation. A syringe was also integrated so that fluidic flow in the cartridge could be controlled by syringe-induced aspiration, thereby eliminating the need for external pumps.Fig. 4Enhancing user-friendliness of MONITOR via a self-contained and sequence-controllable microfluidic cartridge.a Overview of the MONITOR cartridge consisting of four sample storage chambers, one reaction chamber, and one detection chamber. A rotary valve was designed to achieve reaction sequence control, and a syringe was designed to provide the driving force. b Distribution of preloaded assay reagents in the storage and reaction chambers. c The explosion diagram that illustrates all 3D-printed components and assembly protocol for fabricating the MONITOR cartridge. d Photos to demonstrate the control of liquid flow within the MONITOR cartridge via the rotary valve using dye solutions of different colors. e Kinetic curves for the in-cartridge detection of synthetic BKV DNA standard at concentrations 3 × 103 cp/mL to 3 × 107 cp/mL in the original samples using MONITOR. Data are displayed as mean ± SD (n = 3). f Evaluating the LOD of MONITOR for analyzing low concentrations of synthetic BKV DNA standards from 6 × 102 cp/mL to 2 × 103 cp/mL. Data are displayed as mean ± SD (n = 20). Each error bar represents one standard deviation from 20 independent experiments. Rates of positive results were then analyzed to determine the LOD of MONITOR. g Specificity of MONITOR for BKV against varying viral DNA. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. Statistical comparisons were performed using one-way ANOVA followed by a two-sided Dunnett’s multiple-comparisons test, with NTC as the common control. Multiplicity-adjusted P values are reported. ****P < 0.0001; ns, P > 0.05. Source data are provided as a Source Data file. Enhancing user-friendliness of MONITOR via a self-contained and sequence-controllable microfluidic cartridge. a Overview of the MONITOR cartridge consisting of four sample storage chambers, one reaction chamber, and one detection chamber. A rotary valve was designed to achieve reaction sequence control, and a syringe was designed to provide the driving force. b Distribution of preloaded assay reagents in the storage and reaction chambers. c The explosion diagram that illustrates all 3D-printed components and assembly protocol for fabricating the MONITOR cartridge. d Photos to demonstrate the control of liquid flow within the MONITOR cartridge via the rotary valve using dye solutions of different colors. e Kinetic curves for the in-cartridge detection of synthetic BKV DNA standard at concentrations 3 × 103 cp/mL to 3 × 107 cp/mL in the original samples using MONITOR. Data are displayed as mean ± SD (n = 3). f Evaluating the LOD of MONITOR for analyzing low concentrations of synthetic BKV DNA standards from 6 × 102 cp/mL to 2 × 103 cp/mL. Data are displayed as mean ± SD (n = 20). Each error bar represents one standard deviation from 20 independent experiments. Rates of positive results were then analyzed to determine the LOD of MONITOR. g Specificity of MONITOR for BKV against varying viral DNA. Data are displayed as mean ± SD (n = 3). Each error bar represents one standard deviation from three independent experiments. Statistical comparisons were performed using one-way ANOVA followed by a two-sided Dunnett’s multiple-comparisons test, with NTC as the common control. Multiplicity-adjusted P values are reported. ****P < 0.0001; ns, P > 0.05. Source data are provided as a Source Data file. To achieve rapid prototyping, we employed stereolithography 3D printing technique for fabricating the MONITOR cartridge (Fig. 4c and Supplementary Fig. 19). The main body and the syringe were fabricated using transparent polymethyl methacrylate resins. The rotary valve was fabricated using two types of resins via a print-pause-print procedure. While the top layer of the valve was made of the regular polymethyl methacrylate resin, the bottom layer was built using a flexible thermoplastic polyurethane resin to ensure leakage-free packaging. Moreover, the channel within the valve was designed in a rectangular U-shape, so that all connections were made vertically (Supplementary Fig. 20). Upon packaging using a screw-type fastener, our valve design could withstand a burst pressure of over 1600 mbar, whereas previously reported linear valve or single material designs could only sustain up to 100 mbar (Supplementary Fig. 21). The rectangular U-shape channel was also found to promote reagent mixing upon passing through the valve, thereby improving efficiencies of downstream incubation steps in the reaction chamber. Once fabricated and packaged, we further coated the channels and chambers of the cartridge using 2% PEG8000 via a passive adsorption and heat curing procedure at 60 °C, allowing effective reduction of nonspecific adsorption of assay reagents (Supplementary Fig. 22). The functionality of the MONITOR cartridge was initially assessed using four food-colored dyes: yellow to denote the urine sample mixed with the lysis buffer, blue for the RPA reagents, green for helper strands, and red for the gDz sensor (Fig. 4d). The process commenced by rotating the valve from 0° to 36° and slightly drew the syringe to provide the driving force for the yellow dye. Blue, green, and red dyes were sequentially added and blended in the reaction chamber by positioning the rotary valve at 72°, 108°, 144°, respectively. The reaction mixture was then sealed in the reaction chamber by rotating the valve to 180° and then transferred into the detection chamber by further drawing the syringe. Photos in Fig. 4d showed that the sequence of reagent addition was precisely controlled in the MONITOR cartridge with no leakage or nonspecific adsorption. Having validated the functionality of the cartridge, we evaluated the overall performance of MONITOR platform using synthetic BKV DNA. Kinetic curves in Fig. 4e revealed that MONITOR was quantitative and highly sensitive with samples containing 3 × 103 cp/mL BKV DNA still distinguishable from the blank. To further estimate the LOD, we performed 20 replicate detections on ultra-low concentrations of BKV DNA, following the US Food and Drug Administration (FDA) guidelines for in vitro diagnostic device performance. Samples with input concentration above 1 × 103 cp/mL were detected with 100% certainty, while detection probabilities dropped to 85% at 8 × 102 cp/mL and 65% at 6 × 102 cp/mL, respectively. Therefore, the LOD of MONITOR was 1 × 103 cp/mL, representing near single-molecule detection capacity for BKV analysis (Fig. 4f and Supplementary Fig. 23). MONITOR is also highly specific with no cross-reactivity against a panel of commonly encountered viruses, including CMV (Cytomegalovirus), JCV (John Cunningham virus), TTV (Torque teno virus), EBV (Epstein-Barr virus), MCP (Merkel cell polyomavirus), HBV (Hepatitis B virus), and HPV (Human papillomavirus) (Fig. 4g and Supplementary Fig. 24) Enabling on-site quantification via a portable real-time fluorescence reader with a smartphone APP-based reporting system As the clinical outcome of BKV infection depends strongly on the viral load, it is critical to endow MONITOR with on-site quantification capacity. Thus driven, we sought to fabricate a battery-powered portable real-time fluorescence reader (MONITOR reader) that provides more robust quantification capacity over the semi-quantitative lateral flow strip or endpoint fluorescence readouts. To ensure the fidelity of each test, we designed our reader to contain two independent real-time fluorescence detection modules with one for the sample and the other for the negative control. All optical components for each module, including light emitting diodes (LED) at 635 nm, optical lenses, bandpass filters, and photodetectors, were assembled within a compact 37 mm × 32 mm × 24 mm detection module (Fig. 5a and b). A printed circuit board (PCB) equipped with a 16-bit Analog-to-Digital Converter and an ESP32 microcontroller was designed to achieve real-time fluorescence collection and low-power wireless data transmission (Supplementary Fig. 25 and Fig. 26). A smartphone APP was also developed to achieve real-time acquisition and processing of the fluorescence signal and automatic generation of corresponding kinetic curves (Fig. 5c and d).Fig. 5MONITOR reader coupled with smartphone APP for on-site BKV quantification.a Schematic illustration of the MONITOR reader consisting of two real-time fluorescence detection modules, three temperature control zones, a unified control module, and a battery. Bluetooth connection module was employed to achieve wireless connection between the reader and smartphone. b Diagram to illustrate the optical and electrical designs of the MONITOR reader. c User interface of the smartphone APP for controlling the MONITOR reader and for the real-time fluorescence data collection and analysis. Technical triplicates were performed for each sample (n = 3). d Representative kinetic curves for positive and negative samples collected using the MONITOR reader and smartphone APP. The insert images were photos of the detection chamber upon MONITOR at 30 min. e The photo of the MONITOR reader together with a smartphone. Source data are provided as a Source Data file. MONITOR reader coupled with smartphone APP for on-site BKV quantification. a Schematic illustration of the MONITOR reader consisting of two real-time fluorescence detection modules, three temperature control zones, a unified control module, and a battery. Bluetooth connection module was employed to achieve wireless connection between the reader and smartphone. b Diagram to illustrate the optical and electrical designs of the MONITOR reader. c User interface of the smartphone APP for controlling the MONITOR reader and for the real-time fluorescence data collection and analysis. Technical triplicates were performed for each sample (n = 3). d Representative kinetic curves for positive and negative samples collected using the MONITOR reader and smartphone APP. The insert images were photos of the detection chamber upon MONITOR at 30 min. e The photo of the MONITOR reader together with a smartphone. Source data are provided as a Source Data file. To facilitate the MONITOR assay, three temperature zones were also integrated into the MONITOR reader, including two heating zones (95 °C and 37 °C) made of positive temperature coefficient (PTC) heaters and a cooling zone made of a mini-Peltier element to achieve rapid cooling (Supplementary Fig. 27). Aluminum cartridge holders with wedge-shaped structures customed for MONITOR cartridges were designed for all temperature zones to ensure fast thermal conduction, ensuring robust assay performances (Supplementary Fig. 28 and Supplementary Fig. 29). Although containing two detection modules, three temperature zones, and a lithium battery, our reader remained highly compact and lightweight with the size of 150 mm × 165 mm × 130 mm and weight of ~1 kg (Fig. 5e), ensuring high portability and flexibility for uses in remote areas. Assessing the clinical performance of MONITOR We first evaluated the diagnostic performance of the MONITOR assay using 69 clinically confirmed DNA samples extracted from blood and urine of KTRs with or without BKV infection. Standard qPCR test was used to quantify BKV viral loads in all positive samples, which ranged from 1.2 × 10³ to 1.4 × 10¹⁰ cp/mL (Supplementary Fig. 30). The assay correctly identified BKV DNA in 40 of the 40 positive samples and none of the 29 negative samples, representing 100% clinical sensitivity (95% CI, 91.2–100%) and specificity (95% CI, 88.1–100%) (Supplementary Fig. 31). We next assessed the fully integrated MONITOR platform incorporating the microfluidic cartridge and the portable fluorescence reader using 36 midstream urine samples and 15 whole blood samples collected from BKV positive and negative KTRs. BKV viral loads ranged from 1.0 × 10³ to 1.4 × 10¹⁰ cp/mL in positive urine samples and from 1.2 × 104 to 2.6 × 105 cp/mL in positive blood samples (Supplementary Fig. 30). Unprocessed urine samples and 3-time diluted blood samples were directly loaded into the cartridge using disposable plastic Pasteur pipettes. All subsequent test steps were then conducted within the cartridge by sequentially rotating the rotary valve and the test results were automatically collected and analyzed via the smartphone APP (Fig. 6a and detailed in Supplementary Table 1 and Table 2). Results were compared with the standard clinical qPCR assay at the central diagnostic laboratory of West China Hospital of Sichuan University (Fig. 6b).Fig. 6Assessing the clinical performance of MONITOR.a Photos to illustrate the workflow of MONITOR. MONITOR was initiated by loading the urine sample into the cartridge. Urine sample was mixed with the lysis buffer and transferred into the reaction buffer for nucleic acid releasing at 95 °C. RPA reagent was then loaded into the reaction buffer for nucleic acid amplification at 37 °C. Helper strands were then loaded into the reaction chamber, and a snap cooling protocol was employed to expose the detection window in the RPA amplicon. BKV-specific gDz sensor was transferred into the reaction chamber to initiate target recognition and amplified detection via trans-cleavage of fluorogenic reporters. Real-time fluorescence was collected and analyzed in situ via the smartphone APP, enabling on-site reporting of the clinical test results. b Assessing the clinical performance of MONITOR by benchmarking it against standard qPCR-based clinical test. c Kinetic curves of 24 BKV positive and 12 negative KTR urine samples obtained using MONITOR. d Statistical analysis of the 36 urine samples determined using MONITOR with endpoint fluorescence at 30 min. P values were calculated by two-tailed Student’s t-test. The dashed line represents the clinical cut-off determined via ROC analysis (n = 36). e Concordance between MONITOR and qPCR results for 36 urine samples. f Correlation between quantitative test results obtained by MONITOR and qPCR. The solid line represents the best-fit simple linear regression line, and the area between dashed lines represented the 95% confidence interval. Source data are provided as a Source Data file. Assessing the clinical performance of MONITOR. a Photos to illustrate the workflow of MONITOR. MONITOR was initiated by loading the urine sample into the cartridge. Urine sample was mixed with the lysis buffer and transferred into the reaction buffer for nucleic acid releasing at 95 °C. RPA reagent was then loaded into the reaction buffer for nucleic acid amplification at 37 °C. Helper strands were then loaded into the reaction chamber, and a snap cooling protocol was employed to expose the detection window in the RPA amplicon. BKV-specific gDz sensor was transferred into the reaction chamber to initiate target recognition and amplified detection via trans-cleavage of fluorogenic reporters. Real-time fluorescence was collected and analyzed in situ via the smartphone APP, enabling on-site reporting of the clinical test results. b Assessing the clinical performance of MONITOR by benchmarking it against standard qPCR-based clinical test. c Kinetic curves of 24 BKV positive and 12 negative KTR urine samples obtained using MONITOR. d Statistical analysis of the 36 urine samples determined using MONITOR with endpoint fluorescence at 30 min. P values were calculated by two-tailed Student’s t-test. The dashed line represents the clinical cut-off determined via ROC analysis (n = 36). e Concordance between MONITOR and qPCR results for 36 urine samples. f Correlation between quantitative test results obtained by MONITOR and qPCR. The solid line represents the best-fit simple linear regression line, and the area between dashed lines represented the 95% confidence interval. Source data are provided as a Source Data file. Kinetic measurements from MONITOR revealed clear discrimination between BKV positive (n = 24) and negative (n = 12) urine samples (p < 0.0001) (Fig. 6c). A cutoff value for the 30-min endpoint fluorescence signals was established using receiver operating characteristic (ROC) curve analysis (Fig. 6d and Supplementary Fig. 32). Benchmarking against the reference qPCR assay, MONITOR demonstrated 100% clinical sensitivity (95% CI, 85.8–100%) and specificity (95% CI, 73.5–100%) (Fig. 6e). Moreover, MONITOR demonstrated strong quantitative correlation with qPCR (Pearson’s r = 0.8642), accurately quantifying viral loads across a range of 103 to 1010 cp/mL (Fig. 6f and Supplementary Fig. 33). This quantitative capacity allows for the identification of not only BKV-positive patients but also high-risk KTRs requiring active surveillance and prompt clinical intervention. Using a normalized MONITOR signal threshold of 5, our platform identified high-risk KTRs with 92.3% (12 out of 13) concordance to standard qPCR using the clinical threshold of 105 cp/mL. Like urine sample analysis, MONITOR also demonstrated high detection sensitivity for synthetic BKV DNA in whole blood with an estimated LOD at 3 × 103 cp/mL (Supplementary Fig. 34). Further clinical validation using 15 whole blood samples also demonstrated 100% clinical sensitivity (95% CI, 59.0–100%) and specificity (95% CI, 63.1–100%), as well as high quantification consistence with qPCR for whole blood BKV analysis using the fully integrated MONITOR platform (Supplementary Fig. 35). Field trial of MONITOR for opportunistic BKV infection post-transplantation Having established clinical performance of MONITOR under laboratory conditions, we next deployed it for the on-site BKV screening in a resource-limited field setting devoid of laboratory infrastructure (Fig. 7a). The field trial was conducted in Dazhou, a city located over 400 km away from the West China Hospital of Sichuan University in Chengdu. In Dazhou, where no local BKV testing is available, KTRs must routinely travel over five hours by road to Chengdu for opportunistic infection screening (Fig. 7b).Fig. 7A field-trial to validate MONITOR for the on-site monitoring of opportunistic infections in resource-limited settings.a Schematic illustration of the workflow for the on-site urine sample collection and testing. b The map to show the location and travel distance for the field-trial (Made with Natural Earth). c Photos to show the environment for conducting the field trial, where no laboratory facility was available. d A photo showcasing the MONITOR platform, including lyophilized reagents and cartridge sealed into aluminum plastic packaging bags, a portable reader, and a backup battery. e Kinetic curves derived from MONITOR platform for 21 KTR urine samples collected on-site. f Positive and negative test results of the 21 urine samples, with 8 low-risk and 3 high-risk BKV positive infections were identified. Cut-off values were pre-determined by the laboratory assessment of MONITOR. g Concordance between MONITOR and qPCR testing for 21 on-site collected urine samples. h Quantitative correlation between MONITOR and qPCR. The solid line represents the best-fit simple linear regression line, and the area between dashed lines represented 95% confidence interval. i MONITOR versus subsequent qPCR analysis for confirmation of BKV in urine and blood from two initial urine screened-positive KTRs. Source data are provided as a Source Data file. A field-trial to validate MONITOR for the on-site monitoring of opportunistic infections in resource-limited settings. a Schematic illustration of the workflow for the on-site urine sample collection and testing. b The map to show the location and travel distance for the field-trial (Made with Natural Earth). c Photos to show the environment for conducting the field trial, where no laboratory facility was available. d A photo showcasing the MONITOR platform, including lyophilized reagents and cartridge sealed into aluminum plastic packaging bags, a portable reader, and a backup battery. e Kinetic curves derived from MONITOR platform for 21 KTR urine samples collected on-site. f Positive and negative test results of the 21 urine samples, with 8 low-risk and 3 high-risk BKV positive infections were identified. Cut-off values were pre-determined by the laboratory assessment of MONITOR. g Concordance between MONITOR and qPCR testing for 21 on-site collected urine samples. h Quantitative correlation between MONITOR and qPCR. The solid line represents the best-fit simple linear regression line, and the area between dashed lines represented 95% confidence interval. i MONITOR versus subsequent qPCR analysis for confirmation of BKV in urine and blood from two initial urine screened-positive KTRs. Source data are provided as a Source Data file. We performed a double-blind field trial involving 21 locally recruited KTRs. Urine samples were collected and analyzed immediately on-site using MONITOR (Fig. 7c). To facilitate field use, reagents for individual test were lyophilized and packaged together with the microfluidic cartridge in vacuum-sealed aluminum-plastic pouches (Fig. 7d). During on-site testing, lyophilized reagents were reconstituted using bottled purified water, preloaded into the MONITOR cartridge, followed by urine sample introduction. To rule out false-positive test results from cross-contamination, each urine sample was processed alongside a negative control (water substituted for urine). All test results were obtained on-site within one hour. Residual urine samples were shipped to the diagnostic laboratory of the West China Hospital of Sichuan University for standard qPCR benchmarking. MONITOR identified 11 out of the 21 samples as BKV positive during on-site testing (Fig. 7e and f), showing complete qualitative and quantitative concordance with the subsequent qPCR analyses (Fig. 7g and h, Supplementary Figs. 36 and 37). On-site quantitative analysis via the smartphone APP further flagged three samples exceeding the predefined risk threshold (normalized MONITOR signal > 5) (Fig. 7h), a finding later confirmed by subsequent laboratory qPCR testing. Upon receiving their positive on-site results, two KTRs (patient #3 and #20) underwent further clinical follow-up with both urine and blood testing (Fig. 7i). Patient #20 initially screened as BKV positive but low-risk (signal at 2.48), was confirmed to have localized urinary BKV infection. In contrast, Patient #3, initially identified as high-risk (signal at 5.83), was diagnosed with systemic BKV infection, requiring immediate clinical intervention. Together, these results demonstrate that the MONITOR platform enables rapid, accurate, and quantitative on-site screening, supporting early identification of high-risk KTRs and facilitating timely clinical management. Discussion MONITOR is a fully integrated, sample-in-answer-out assay system for rapid monitoring opportunistic infections post transplantation from unprocessed urine samples in nonlaboratory settings. Its development encompasses systematic innovation spanning diagnostic reagent, microfluidic cartridge, and instrumentation. At the core of the assay is a gDz-based, CRISPR-like mechanism, which achieves near single-molecule sensitivity upon coupling with RPA while offering substantially greater reagent stability than conventional CRISPR-based diagnostics. Upon lyophilization, MONITOR reagents remain stable for over one year at ambient temperature and can be transported without cold-chain requirements over distances exceeding 4000 km. The improved HUDSON protocol allows one-step, extraction-free sample processing in just 3 min. Operation is further simplified by a self-contained, sequence-controllable microfluidic cartridge, allowing the test to be performed reliably by users without specialized training (user instruction is provided in Supplementary Table 1 and Table 2). Quantitative detection and on-site data analysis are also accomplished using a portable, battery-powered fluorescence reader coupled with a smartphone APP. Collectively, these innovations render the MONITOR platform highly aligned with the ASSURED (affordable, sensitive, specific, user-friendly, rapid, equipment-free, delivered) criteria that are widely accepted as the benchmark for an ideal POCT43. Furthermore, the smartphone-enabled real-time connectivity and minimal specimen processing requirements further adapt the MONITOR system to the evolving REASSURED criteria for next-generation POC diagnostics43. These features establish MONITOR a versatile and scalable platform with broad applicability for POCT, as demonstrated by its use in monitoring post-transplant opportunistic infections. Our findings demonstrate that MONITOR provides an effective solution for POC monitoring of BKV infection in KTRs, enabling frequent surveillance that is critical for preventing allograft loss. Persistent BKV infection, which develops in 10% to 12% of KTRs1,2,6,44,45, can progress to BKV-associated nephropathy (BKVAN), a condition leading to irreversible graft dysfunction in > 90% and graft loss in > 50% of affected individuals3. In our clinical cohort, among the 71 BKV-positive patients, 32 were classified as having different levels of BKPyVAN risk, and 2 had biopsy-confirmed BKPyVAN (Supplementary Table 3). As BKV infection is largely asymptomatic, regular screening with quantitative POC testing is critically needed, particularly in the first two years after kidney transplantation8. Moreover, in KTRs requiring increased immunosuppression or antirejection therapy, monthly screening for BKV infection for next 3 months is also recommended8. While methods, such as RPA-CRISPR techniques and loop-mediated isothermal amplification (LAMP) have been developed for rapid BKV detection, they often require separate DNA extraction, multi-step operations, and yield only qualitative results, limiting their use for viral load quantification in resource-limited settings30,34,36,46. In contrast, MONITOR platform integrates sample lysis via brief thermal incubation, nucleic acid amplification, and real-time fluorescence detection within a sealed microfluidic cartridge, eliminating cross-contamination risks and user intervention. This system achieved 100% concordance with laboratory qPCR in both controlled and field settings, and successfully identified KTRs with high-risk systemic infections during an on-site trial, underscoring its clinical utility for guiding timely therapeutic management. Based on the exciting findings in this study, we plan to conduct multicenter clinical validation with larger sample size in the near future to facilitate the clinical translation of MONITOR technology. Further technical innovations should be made to enable MONITOR for use in any location, such as at-home testing. Toward this goal, we aim to innovate sample processing and signal generation protocols, so that MONITOR can take place in ambient temperature without the need for high-temperature heating. We also plan to upgrade the MONITOR cartridge by equipping it with an automated control module, so that the entire testing process can be completed with the push of a single button. To enable at-home testing, we will also incorporate MONITOR with internal controls, such as sensors for patients’ genomic DNA, to minimize false-negative diagnostic results. The combination of internal control for false negatives and the current external control for false positives will greatly enhance test fidelity. Together, these improvements could keep enhancing the accuracy, accessibility, and user-friendliness of MONITOR, positioning it as a versatile diagnostic tool in the fight against opportunistic infections and diverse healthcare threats in immunecompromised populations. Methods Ethical statement This study was approved by the Ethics Committee of West China Hospital, Sichuan University (No. 20231193). Informed written consents were obtained from all participants. All research was carried out under relevant guidelines and regulations. Study design This study was designed to evaluate the performance of the MONITOR platform for the detection of BKV DNA in unprocessed urine samples from KTRs. We investigated the sensitivity, specificity, and quantitative capabilities of the MONITOR assay, as well as its practicality and effectiveness in a real-world, resource-limited setting. Experiments were conducted using clinical urine samples from KTRs, lyophilized reagents, and a fully integrated microfluidic cartridge. In vitro assays (sample processing, nucleic acid amplification, and signal detection) were performed using the MONITOR platform. Mechanistic studies integrated gDz screening, chemical modifications, and kinetic analysis. Clinical validation of single-agent and combination detection methods was tested using 120 clinical samples from KTRs, and effects on the detection of viral DNA were assessed using standard qPCR-based tests. For the field trial, sample sizes were determined based on the availability of KTRs in Dazhou City and the need to ensure a diverse representation of patients. Urine samples were collected from 21 KTRs. All samples were included in analyses, unless they met predefined exclusion criteria (insufficient volume, contamination, or incomplete metadata). Sample sizes and replicates are provided in the corresponding figure legends. See the Supplementary Materials for complete materials and methods. Materials All DNA oligonucleotides (Supplementary Table 4) were synthesized and modified by Sangon Biotech Co., Ltd. (Shanghai, China), and were purified by high-performance liquid chromatography (HPLC). Magnesium chloride hexahydrate (MgCl2 ∙ 6H2O), 2 M Tris-HCl, 1 × Tris EDTA (TE) buffer, trehalose, Tween 20, and RNase-free water were purchased from Sangon Biotech Co., Ltd. (Shanghai, China). 2×TBE-Urea loading buffer were purchased from Zhongke Ruitai Biotechnology Co., Ltd. (Beijing, China). TwistAmp® Basic kit for RPA was purchased from Twist-Dx. RNaseAlert® Lab Test Kit v2 was purchased from Thermo Fisher Scientific. ClearV4 and Elastic 50 A resin were purchased from Formlabs. Premium ABS (1.75 mm) wires were purchased from Fuzhi Information Technology Co., Ltd. (Shanghai, China). Ultrapure (UP) H2O (18.25 MΩ*cm) was purified using an ULUPURE ultrapure water system (Chengdu, China). Screening of BKV-specific gDz A 10–23 RNA-cleaving Dz was chosen as the template for screening BKV-specific gDz. The panel of arm sequences were designed according to a highly conserved domain within the tAg coding region of BKV and then screened via PAGE analysis. For each test, 0.5 μM gDz was mixed with 2.5 μM Cy5-labeled substrate (also referred to as reporter) in 10 mM Tris-HCl (pH = 8.0). Each reaction was initiated by adding 1 μL MgCl₂ solution at a final Mg2+ concentration at 1 mM and quenched by adding 2× TBE-urea loading buffer, followed by heating at 95 °C for 10 min and rapid cooling. Samples were loaded onto a 15% denaturing polyacrylamide gel containing 8 M urea. Electrophoresis was performed using a Bio-Rad system at 120 V in 1× TBE buffer until complete separation of substrate and DNAzyme bands. Gels were imaged using a GenSens 2200 system and analyzed semi-quantitatively via ImageJ. Apparent rate constants (kobs) were calculated by nonlinear fitting of cleavage percentages over time using the pseudo-first-order kinetic equation:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${Y}_{t}={Y}_{\max }\times \left(1-{e}^{-{k}_{{obs}}t}\right)$$\end{document}Yt=Ymax×1−e−kobstwhere Yt is the cleavage yield at time t, and the Ymax is the maximum yield. Detection of BKV DNA using gDz sensor Once obtaining the optimal BKV-specific gDz, it was engineered into a gDz sensor for detecting BKV DNA by blocking the arm sequences using a complementary sequence. Upon the detection of BKV DNA, a typical reaction containing 20 nM gDz sensor, 100 nM Cy5-BHQ3 dual-labeled reporter, and varying concentrations of target DNA in a reaction buffer (10 mM Tris-HCl, 50 mM MgCl₂, 0.1% Tween 20, pH 8.0) was incubated at 37 °C. The fluorescence was measured immediately upon preparing the reaction mixture with a data acquisition rate of one read per minute for 100 min using a BioTek Cytation 5 cell imaging multimode microplate reader (Agilent Technologies, Inc., USA) with excitation at 649 nm and emission at 671 nm. MONITOR assay protocol Primers were adopted from the previous study by Collins and colleagues29, which amplified tAg coding region of BKV containing a 26 nt target sequence for the gDz sensor. For a typical MONITOR assay, RPA was first conducted by mixing 2.5 μL sample with 15 μL rehydrated RPA mixture, 2.5 μL of 10 μM forward and reverse primers, and 7.5 μL nuclease-free water. Amplification was initiated by adding 2.5 μL 280 mM magnesium acetate and incubated at 37 °C for 20 min. After amplification, the RPA reaction mixture was mixed with 20 μL helper strands (500 nM) in a buffer containing 10 mM Tris-HCl, 3 mM MgCl₂, and 0.1% Tween 20, heated at 95 °C for 3 min, and rapidly cooled for 2 min. To this mixture, 40 nM BKV-gDz sensors and 200 nM reporters were added (50 μL), resulting in a final reaction mixture of 97.5 µL. Subsequently, 90 µL of the final reaction mixture was transferred to a microplate. The fluorescence was measured immediately with a data acquisition rate of one read per minute for 100 min using a BioTek Cytation 5 cell imaging multimode microplate reader (Agilent Technologies, Inc., USA) with excitation at 649 nm and emission at 671 nm. Protocol for one-step urine sample processing For a typical MONITOR test with optimal sample processing protocol, the clinical urine sample was mixed with the lysis buffer containing 10 mM TCEP and 1 mM EDTA and then heated at 95 °C for 3 min. To evaluate the RNase activity upon sample processing, 45 μL treated urine sample was mixed with 5 μL RNaseAlertv2 substrates and incubated at 37 °C. The fluorescence was measured immediately with a data acquisition rate of one read per minute for 60 min using a BioTek Cytation 5 cell imaging multimode microplate reader (Agilent Technologies, Inc., USA) with excitation at 485 nm and emission at 520 nm. Lyophilization of MONITOR assay reagents To prepare lyophilized gDz sensor, a mixture containing 40 nM gDz sensor and 200 nM reporter in 50 μL lyophilization buffer containing 1% trehalose was frozen at −80 °C in 200 µL Eppendorf tube. Then, the frozen reagents were lyophilized for 4 h in freeze dryer (LabogeneTM, Gene Company Limited). Helper strands were lyophilized separately using the same lyophilization protocol. For determining the long-term stability of lyophilized assay reagents at ambient temperature, lyophilized reagents were dissolved at a given storage timepoint, the performance of which was then evaluated by detecting 6 × 107 cp/mL synthetic BKV DNA standard using the MONITOR assay protocol as outline above. Fabrication of MONITOR cartridge The MONITOR cartridge was designed using Fusion360 software (Autodesk) and fabricated using a stereolithography 3D printer (Formlabs). ClearV4 resin was used to 3D print the main body of the cartridge and the syringe. Ultraviolet (UV) exposure at 25 mJ/cm2 with 25-μm-thick layers was set for ClearV4 resin and the post-cured were performed under UV light (405 nm) at 60 °C for 10 min. The valve was printed using both ClearV4 resin and the flexible Elastic 50 A resin via a print-pause-print protocol47. Specifically, the top layer of the rotary valve was first printed using ClearV4 resin using the same parameter setup as for the main body of the cartridge. Once printing the top layer, surface was cleaned to remove any uncured resin and then started the printing again by replacing the ClearV4 resin with the flexible resin and set the UV exposure at 50 mJ/cm2 with 100-μm-thick layers. After printing, the cartridge components were cleaned in a post-processing device using isopropanol to remove uncured resin from both surfaces and internal channels. These components were then post-cured in UP water (70 °C) for 5 min. The main body and the valve were then assembled via a steel fastener. The reaction chamber and the detection chamber were finally sealed using delayed adhesive microfluidic tape (9795 T, 3 M). Development of the MONITOR reader and the smartphone APP The MONITOR reader was designed using Fusion360 software (Autodesk) and the main body was fabricated using 3D printing (Hyper Speed ABS Filament, Raise3d Pro2). The optical module comprised a red LED, an objective lens, an excitation band-pass filter, a dichroic mirror, an emission band-pass filter, a condenser lens and a photodiode. An operational amplifier (OP-AMP) was employed to convert the current signal generated by the photodiode into a voltage signal. This analog signal was then digitized through a 16-bit analog-to-digital converter (AD7705). The heating module was crafted from aluminum alloy and tailored to accommodate the MONITOR cartridge. A custom-dimensioned PTC heater (12 V) was used to provide stable heat to the aluminum block. Accurate temperature control was achieved by implementing a proportion integration differentiation (PID) algorithm. The compact cooling module consists of a Peltier element (35*20 mm, 12 V), an aluminum sample holder, heat sink fins, and a fan. The Peltier cooler generates a temperature difference of 69 °C (~24 W of cooling capacity). The sample holder is fixed to the cooling end of the Peltier element to provide the low temperature required for the quenching process. All contact surfaces are covered with thermal grease to ensure effective heat conduction. The system control was achieved via a custom-designed PCB (Shenzhen, JLC). The ESP-IDF framework (v5.2.1) and C language were used to develop the firmware of the ESP32 microcontroller (MCU). Bluetooth Low Energy (BLE) was integrated for wireless communication with smartphones. During wireless communication, the transmitted data were divided into three categories, including instrument control (8 bytes), instrument status (20 bytes), and fluorescence data (16 bytes). The smartphone APP for the MONITOR reader was developed based on JAVA and the Android Software Development Kit (SDK). This APP featured user-friendly graphical user interfaces (GUI) for temperature control, parameter adjustment, real-time fluorescence data collection, and reporting of the test result. Assay protocol with the fully integrated MONITOR platform First, lyophilized reagents were dissolved and preloaded into the storage chambers of the MONITOR cartridge, which was then sealed using custom-shaped adhesive microfluidic tape (9795 R, 3 M). Urine sample (5 µL) was then loaded into the cartridge through the sample inlet using a disposable micropipette. Upon mixing with equal volume of the lysis buffer, the mixture was transferred into the reaction chamber by switching the rotary valve to 36°. Upon heating at 95 °C for 3 min within the heating module of the MONITOR reader, RPA reagents (25 µL) were loaded into the reaction chamber by switching the valve to 72°. The cartridge was inserted into the 37 °C heating zone of the reader for 20 min for RPA. Afterward, the valve was rotated to 108° to add helper strands (20 µL) into the reaction chamber. The snap cooling was then performed by first placing the cartridge at 95°C heating zone for 3 min and then 4°C cooling zone for 2 min. The gDz sensor (50 µL) was finally added into the reaction chamber by further switching the valve to 144° and real-time fluorescence data was then collected by placing the cartridge to the detection module. The temperature control, real-time fluorescence data collection, and data analysis were all performed using the smartphone APP. The costs of fully integrated MONITOR platform were presented in Supplementary Table 5. Clinical validation of MONITOR For in-lab clinical validation of MONITOR, de-identified urine samples collected from KTRs were provided by the West China Hospital (WCH) of Sichuan University. Urine sample analysis was conducted using the assay protocol with the fully integrated MONITOR platform, which was then compared to the clinical reported qPCR-based test results obtained from the core diagnostic facility of WCH. For the field trial, 21 KTRs who located at Dazhou City but conducted their kidney transplant surgeries at WCH were recruited for the on-site BKV testing. Urine samples were collected on-site and then tested using MONITOR with the above-mentioned assay protocol. Figure preparation All schematic illustrations were created by the authors using Adobe Illustrator (version 2020). The 3D models were created and rendered using Autodesk Fusion 360 software. The map of the field-trial locations was generated in QGIS (version 3.44) using public-domain shaded-relief raster data from Natural Earth. Statistics and reproducibility In this study, no statistical method was used to predetermine sample size. For assay development, condition optimization and analytical validation, sample sizes were selected according to the experimental design, and the number and type of replicates are specified in the corresponding figure legends. For clinical validation, sample sizes were determined by the number of eligible samples available. All eligible samples were included in the analyses, and no data were excluded. The experiments were not randomized because this was a diagnostic-accuracy study in which all samples were evaluated against qPCR as the reference standard. In the field validation, all samples were evaluated under blinded conditions, and MONITOR testing and qPCR analysis were conducted independently. Statistical analyses were performed using GraphPad Prism software (version 10.0). Data are presented as the mean ± standard deviation (SD), unless otherwise indicated. For comparisons between two groups, statistical significance was assessed using a two-tailed Student’s t-test. For comparisons involving more than two groups, one-way analysis of variance (ANOVA) followed by Dunnett’s multiple-comparisons test was used. P values of less than 0.05 were considered statistically significant. The sensitivity and specificity of MONITOR were calculated using qPCR results as the reference standard. The cutoff value for the 30 min endpoint fluorescence signal was determined by receiver operating characteristic (ROC) curve analysis. Exact two-sided 95% confidence intervals for sensitivity and specificity were calculated using the Clopper–Pearson method based on the binomial distribution. The correlation and agreement between MONITOR and qPCR measurements were assessed using log10-transformed viral-load values. Agreement between MONITOR and qPCR measurements was evaluated using Bland–Altman analysis by plotting the difference between paired measurements (MONITOR − qPCR) against their mean. The mean difference was used to estimate the bias with 95% limits of agreement. Reporting summary Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article. Supplementary information Supplementary Information Reporting Summary Transparent Peer Review file Supplementary Information Reporting Summary Transparent Peer Review file Source data Source Data Source Data