Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 24.
Published in final edited form as: Biosens Bioelectron. 2022 Aug 24;216:114654. doi: 10.1016/j.bios.2022.114654

MINI: a High-Throughput Point-of-Care Device for Performing Hundreds of Nucleic Acid Tests per Day

Duncan McCloskey 1,*, Juan Boza 1,*, Christopher E Mason 2,3, David Erickson 4,5,†
PMCID: PMC10960951  NIHMSID: NIHMS1970248  PMID: 36084523

Abstract

There are a variety of infectious diseases with a high incidence and mortality in limited resource settings that could benefit from rapid point of care molecular diagnosis. Global health efforts have sought to implement mass-screening programs to provide earlier detection and subsequent treatment in an effort to control transmission and improve health outcomes. However, many of the current diagnostic technologies under development are limited to fewer than 10 samples per run, which inherently restricts the screening throughput of these devices. We have developed a high throughput device called “MINI” that is capable of testing hundreds of samples per day at the point-of-care. MINI can utilize multiple energy sources – electricity, flame, or solar – to perform loop-mediated isothermal amplification (LAMP) in a portable and robust device which is ideal for use in limited resource settings. The unique opto-electronic design of MINI minimizes the energy and space requirements of the device and maximizes the optical isolation and signal clarity, enabling point-of-care analysis of 96 unique samples at once. We show comparable performance to a commercial instrument using two different LAMP assays for Kaposi’s sarcoma-associated herpesvirus and a common housekeeping gene, GAPDH. With a single device capable of running hundreds of samples per day, increased access to modern molecular diagnostics could improve health outcomes for a variety of diseases common in limited resource settings.

Keywords: Molecular diagnosis, screening, point-of-care, loop-mediated isothermal amplification (LAMP), high-throughput

1. Introduction

Molecular diagnostic approaches are powerful tools for screening and diagnosing viral-borne diseases at the point-of-care (POC). POC testing and mass-screening efforts have shown success in quickly identifying illness and enabling rapid implementation of treatment and approaches to reduce transmission. The need for high-throughput diagnostic devices was emphasized by the SARS-COV-2 pandemic that began in 2019(Jiang et al. 2020; Oeschger et al. 2021), however, there are numerous other diseases that would benefit from rapid molecular diagnosis(Chakaya et al. 2021; Detemmerman et al. 2018; Parsons et al. 2011; Sung et al. 2021; Tekola-Ayele and Rotimi 2015; Wu et al. 1989). A high-throughput, POC device could enable large-scale screening in a decentralized format, limiting the burden on traditional healthcare infrastructure and reducing the time to result before treatment is administered(Nakalembe et al. 2020; Nakisige et al. 2020). These benefits would be further evident in limited-resource settings that often do not have access to state-of-the-art laboratories or traditional healthcare facilities.

A gold standard laboratory approach for molecular diagnosis is the quantitative polymerase chain reaction (qPCR) which can be used to measure small amounts of nucleic acids (DNA or RNA) within a sample. These nucleic acids can serve as indicators for disease, either pathogenic or genetic in origin(Nhu et al. 2014; Obande and Singh 2020). However, conventional qPCR methods can be difficult to employ at the POC, particularly in limited-resource settings, due to the complexity of thermal cycling and cost of equipment(Drain et al. 2014; Kozel and Burnham-Marusich 2017; Yager et al. 2008). Loop-mediated isothermal amplification (LAMP) is an alternative approach that offers additional benefits for POC testing, such as single-temperature operation and resistance to inhibitors(Craw and Balachandran 2012; Wei et al. 2022; Zhang et al. 2019). A number of devices have been developed for implementing LAMP testing at the POC for infectious diseases like tuberculosis(Boehme et al. 2011; Nicol et al. 2011), malaria(Morris et al. 2015; Ponce et al. 2017), and COVID-19(Butler et al. 2021; Dao Thi et al. 2020; Panpradist et al. 2021; Yu et al. 2020a; Zhu et al. 2020), as well as for cancers including cervical cancer(Hagiwara et al. 2007; Rohatensky et al. 2018; Yin et al. 2019; Yu et al. 2020b) and Kaposi’s sarcoma(Snodgrass et al. 2018). However, these technologies are lacking the high-throughput capability to analyze potentially hundreds of samples per day, which is necessary for successful mass-screening(Li et al. 2021b; Snodgrass et al. 2018). Some approaches utilize smartphones and inexpensive electronics to perform optical measurements but can only run up to 6 samples at a time(Buultjens et al. 2021; Panpradist et al. 2021; Snodgrass et al. 2018; Wang et al. 2021). Microfluidic chips and paper-based platforms combined with an optical reader have also been developed, though similarly can only process single-digit samples per run(Li et al. 2021a; Liu et al. 2020a; Liu et al. 2020b; Zhang et al. 2021). While some groups claim the ability to multiplex for improved testing capacity, reliance on complicated microfluidic approaches or increasing the size of a device further reduces the accessibility at the point of care. There has been development of a mobile fluorescence reader capable of 96-sample analysis, however, this device does not perform LAMP and can only be used for post-reaction measurements(Hambalek et al. 2021).

In this paper, we introduce the first fully-integrated Multiplexed Isothermal Nucleic acid amplIfication device, MINI, for 96-sample high-throughput molecular diagnosis and screening at the point-of-care. MINI is capable of performing and processing LAMP reactions using a standard 96-well plate in a device the size of a lunchbox. This device features a novel optical arrangement to decrease power consumption, increase resolution, and eliminate cross-interference while analyzing 96 individual wells using three distinct excitation wavelengths. We will first discuss the mechanical features of the MINI device that make it highly accessible for POC use in limited-resource settings. We then elucidate the unique features of MINI that make it possible to analyze 96 samples simultaneously while only requiring a simple laptop with a USB connection for power. We show the consistency of amplification and analysis in MINI by testing positive and negative controls in completely full 96-well plates. Finally, we test MINI using two different LAMP assays and compare performance to a commercial qPCR device. To our knowledge, this is the first portable, high-throughput device capable of analyzing 96 independent samples using LAMP for POC diagnosis and screening.

2. Materials and Methods

2.1. MINI device – mechanical components and construction

Aluminum pieces were fabricated out of 6061-T aluminum by the Cornell Machine Shop. The aluminum outer case of MINI was designed in-house at Cornell and manufactured by ProtoCase. Temperature storage and stabilization is accomplished using a phase-change material specific to 65°C (OM65P-5X6–100, SavENRG), with 250 grams of material melted in a beaker and added to the outer basin of MINI. Heating via flame was performed using a portable butane burner (MTAMT-30, Master) while the MINI bottom panel was removed, and the device elevated 4” using two aluminum blocks. Insulation within the device is lightweight melamine foam insulation sheets 1” thick (9249K24) obtained from McMaster-Carr and trimmed to fit with a knife. All minor hardware and fasteners were also sourced from McMaster-Carr. The optical filter within the device was custom ordered through Omega Optical as a 4” by 3” 530/660nm dual-bandpass filter. Cost breakdown for the MINI can be found in the supplementary information [Supplemental Table 1].

2.2. MINI device – electrical components and construction

The three PCBs (Main, Top, Bottom) were designed using Autodesk – Eagle software and manufactured overseas at PCBWay. All the components were soldered in-house to supervise the electrical performance as the components were added. Two electrical subsystems are controlled by a Teensy 3.6 microcontroller – optical and temperature. The optical subsystem comprises of two 8-bit shift registers (MM74HC595SJ), 24 RYB LEDs (SMTL4-SRYB), and 96 light to frequency converters (TSL237T) purchased from DigiKey. The temperature control subsystem has 3 temperature sensors (2 AD22103KRZ, 1 TMP512), 1 K-type thermocouple(Omega size #8 Screw size), and 1 cartridge heater (MCH1–240W-004, COMSTAT). All electrical components are powered via USB laptop connection with a micro-USB cable type B except the cartridge heater, which uses a separate 12V/6A power supply (SGA60U12-P1J).

2.3. Electrical operation of MINI

MINI temperature control and data acquisition are performed using an on-board microcontroller, while data storage and analysis is accomplished using a custom software capable of running on any standard laptop. The operation of the optical subsystem consists of cycling through the LED wavelengths while taking light measurements. The shift registers turn photodiode columns ON/OFF, enabling data acquisition with only eight analog pins. Before the final design, subsystems were tested separately using breadboards and off-the-shelf components. Temperature control is achieved through a custom algorithm using the outer basin and sample temperature measurements along with the cartridge heater, ensuring component integrity during rapid heating.

2.4. Custom Software – Post-processing and threshold time calculation

The custom software was written using Python 3.7. After data acquisition, the software processes the curves by first running a Hampel filter (window = 11, standard dev. = 3) followed by taking the first discrete difference of element (fluorescent difference). A second Hampel filter is then applied (window = 5, standard dev. = 1) to remove any random voltage drop along with a rolling mean (window = 10) to smooth the data. Reconstruction of the curve is then obtained through the cumulative sum.

Threshold calculation is done by applying a threshold of 0.03 on the fluorescent difference curves and tracing this point back to its respective time. In the case where this threshold is not surpassed (no amplification), the threshold time is set to end of the run.

2.5. LAMP assay for KSHV

Six primers for target Orf26 of KSHV(Kuhara et al. 2007) [Supplemental Table 2] synthesized by IDT were added to a mastermix consisting of isothermal amplification buffer, dNTP mix, MgSO4, nuclease-free water, and Bst 2.0 WarmStart polymerase, all from New England Biolabs [Supplemental Table 3]. Each well of the 96-well plate (AB-0600, Thermo Scientific) received 35μL of mastermix and 5μL of DNA sample. Each sample well was topped with 50μL of laboratory-grade mineral oil to prevent evaporation during the run. Stocks of KSHV plasmid DNA samples were utilized from a previous publication(Snodgrass et al. 2018) to create the positive and negative controls, as well as generate the standard curves for both MINI and the commercial device.

2.6. LAMP assay for GAPDH

Six primers for the housekeeping gene of GAPDH [Supplemental Table 4] were added to a mastermix of isothermal amplification buffer, MgSO4, dNTP mix, Bst 2.0 WarmStart polymerase, and nuclease-free water (all from New England Biolabs). The exact reagent composition of the assay can be found in the supplemental information [Supplemental Table 5]. Each sample well of the 96-well plate received 45μL of mastermix and 5μL of DNA sample, after which each sample well was topped with 50μL of mineral oil to prevent evaporation during the run. GAPDH DNA samples were produced using the QIAGEN DNeasy Blood and Tissue kit (69504, QIAGEN) using tissue samples received from Dr. Ethel Cesarman’s lab at Weill Cornell Medicine. Overnight extraction was completed prior to the purification protocol from QIAGEN followed exactly per manufacturer’s recommendations. Extracted DNA concentration was measured using a SpectraMax QuickDrop micro-volume spectrophotometer using 1.0μL of sample. GAPDH copy number was then estimated using the weight of the human genome(Biosystems 2003) and dilutions were performed using the AE elution buffer from the QIAGEN DNeasy kit.

2.7. Isothermal amplification using the QuantStudio 7

The normal temperature-cycling profile was replaced with a single ramp from room temperature to 65°C, followed by 150 repeated holds at 65°C to enable a total assay time of 50 minutes with measurements taken every 20 seconds, closely mimicking the MINI device. Samples run on the QuantStudio 7 did not have mineral oil added as there is a heated top plate to prevent evaporation.

3. Results and Discussion

3.1. Instrument design and heating characteristics

MINI was developed to serve as a high-throughput point-of-care device for isothermal nucleic acid testing. The device is easily portable to various testing sites with dimensions around the size of a standard lunchbox – 12 inches long by 8 inches wide by 5 inches high – and weighing under 10 pounds fully assembled [Figure 1A]. Two nested rectangular basins – inner and outer – are used to hold the optical system and the phase change material (PCM), respectively [Figure 1B]. The design and construction utilizing two nested aluminum basins makes it possible to exchange the PCM with one that melts at a different temperature, providing flexibility for different assay requirements.

Figure 1 – MINI mechanical and heating properties.

Figure 1 –

Open in a new tab

(A)MINI is 12 inches long by 8 inches wide by 5 inches high and weighs less than 10 lbs. (B) Cross-sectional view shows the two nested aluminum basins (inner and outer) with PCM between. Sample measurement system is contained within the inner basin. (C) Heating via electricity-powered cartridge heater takes ~72 minutes to reach the desired temperature of 65°C to fully melt the PCM. (D) Once MINI is at operating temperature, an inserted fully-loaded 96-well plate reaches reaction temperature within 6 minutes. Without any additional energy input and independent of initial heat source, MINI maintains an average temperature of 65°C for over 90 minutes. At 108 minutes, there is a difference of 2.5°C between the central and corner sample wells. (E) Heating via flame is accomplished with 23 minutes of applied flame and stabilizes at reaction temperature in 36 minutes total. (F) Heating using a Fresnel lens (4ft2) projected onto a black, aluminum top plate is accomplished in ~68 minutes.

An inset cartridge heater in the outer basin is used to heat the device when electricity is available, and heat sinks added to the outer basin interior to allow for faster heat transfer into the PCM. The PCM within the MINI has a long thermal dwell at 65°C as it transitions from solid to liquid, storing heat energy as it melts completely in 72 minutes [Figure 1C]. This enables MINI to continue to maintain temperature within 1°C for over 90 minutes without any additional energy input [Figure 1D]. There is only a slight 2.5°C temperature difference between wells in the center of the array and wells on the periphery after almost two hours of cooling. There is effectively no temperature difference between central and peripheral wells while electric power is present. Temperature uniformity across the 96 sample wells indicates that there should be no difference in amplification activity between samples. While heating by electricity was accomplished in just over 70 minutes, exchanging our current cartridge heater for a higher-wattage equivalent could produce faster heating times.

MINI is also capable of heating using alternate energy sources when electricity is not present, such as flame or solar energy. The bottom of the device is removable, allowing a small butane burner to heat the device in ~33 minutes [Figure 1E]. For solar heating, an aluminum plate painted black is rested on the walls of the outer basin and a large Fresnel lens (4ft2) was attached to focus solar rays in the center of the plate. Using this setup, we were able to fully melt the PCM in just over an hour on a sunny day in New York [Figure 1F]. The test began at 12:00 noon with an air temperature of 19°C and an intermittent 10–14kph breeze. However, even with successful solar heating, the large Fresnel lens does somewhat reduce the portability and simplicity of the device compared to a small flame source such as a butane-powered burner.

3.2. Optical and electrical design to enable 96 sample capacity

The optical system serves to illuminate the 96 sample wells and collect the resulting data, which can include fluorescence, colorimetric, or turbidity measurements. A top printed circuit board (PCB) holds 24 light emitting diodes (LEDs) that illuminate four adjacent sample wells, where light then continues out of the bottom of the sample, through an optical filter, and is collected by photodiodes on a second PCB below [Figure 2A]. This compact LED arrangement allows MINI to utilize readily available 96-well plates similar to commercially available machines.

Figure 2 – MINI optical and electronic properties.

Figure 2 –

Open in a new tab

(A) Light emitted from LEDs on the top PCB travels down and splits to illuminate four adjacent sample wells. Light continues out of the bottom of each sample well, through the optical filter, and is collected by photodiodes below. (B) Each sample well is connected to the LED using individual channels created with an end-mill. The channels are used to completely isolate each sample well from every adjacent sample well. (C) Different wavelengths of light are used to illuminate the samples and measure different emission spectra for fluorescence, normalization, and turbidity. (D) Each column is analyzed over 200ms, requiring 2.4 seconds to complete all 96 measurements for a single wavelength. Additional time added for LED stabilization and temperature measurements brings the total collection time for each wavelength to 5 seconds, and a total data collection cycle time of 15 seconds.

In order to optically isolate adjacent sample wells, as well as minimize the amount of electronics needed – using only 24 LEDs for 96 total samples – MINI utilizes a unique machined design. Individual channels were created within the aluminum sample block using an end-mill to allow light from each LED to illuminate the four adjacent sample wells [Figure 2B]. Each sample well is only connected to a specific LED, without any connections to nearby LEDs or samples, so there is minimal optical noise between samples. Further noise reduction was achieved by increasing the light path length with a black plastic spacer beneath the milled aluminum piece, minimizing any reflected light from the shiny metal above. With this design, we are able to (1) fully isolate each sample from the influence of an adjacent well, and (2) minimize the electronic requirements of the device, allowing us to use only 24 LEDs for illumination. This arrangement also minimizes the overall size of the device, where LEDs are inset between the sample wells instead of above like in a traditional analyzer.

Each of the 24 LEDs cycles from blue to yellow to red, enabling measurement of multiple parameters [Figure 2C]. Blue light can be used to measure fluorescence of Evagreen dye, yellow light is used for a normalization dye such as ROX, and red light can be used to measure the turbidity of the reaction. The dual bandpass filter that MINI uses only allows green (530nm) and red (660nm) wavelengths to pass, further increasing the quality of our signal. Due to the large number of wells and limited number of analog inputs on the microcontroller, MINI employs two 8-bit shift registers that switch on a single column of wells at a time. Data for each column is collected over a 200ms span before moving to the next column [Figure 2D]. In total, 2400ms or 2.4 seconds are used to acquire data from all 96 wells for a single LED wavelength. Each measurement cycle also includes additional time to allow for LED stabilization and temperature recordings, bringing the total time to 5 seconds for each wavelength and 15 seconds for a complete data acquisition cycle. We also have the ability to expand the optical (LED and filter) components for other excitation and emission wavelengths, increasing the number of assays MINI can run.

After a successful run, MINI generates 288 light intensity curves (96 wells and 3 wavelengths) which are analyzed by custom Python software. Data processing consists of a series of Hampel filters to eliminate noise and correct for potential voltage drops. The software then determines the amplification threshold time (in minutes) or determines there was no amplification.

3.3. Determining performance of optical isolation design using positive and negative controls

In order to validate the novel optical isolation design and repeatability between runs across all 96 sample wells, we utilized a LAMP assay for Kaposi’s sarcoma associated herpesvirus (KSHV). We ran two separate plates with 79 positive (PC) and 17 negative (NC) controls arranged to satisfy many possible testing combinations – one NC per row/column, differing number of adjacent PC, etc. [Figure 3]. Using positive control samples with expected strong amplification curves, we analyzed various conditions of potential sample interference with varied number of negative controls adjacent. This testing was done over multiple runs to also assess the repeatability of measurement for the MINI device.

Figure 3 – MINI stability and repeatability across 96 sample wells.

Figure 3 –

Open in a new tab

Positive and negative controls for a KSHV LAMP assay were run in differing configurations in the MINI to assess the performance of our unique optical design and repeatability between runs. Consistent standard deviation of positive control amplification, as well as consistent non-amplification of the negative controls, indicates that each sample well is fully isolated from all adjacent wells.

Overall standard deviation between PC threshold times was 50 seconds, showing consistent results across all 96 wells of the device. We saw strong amplification in all PC samples and no amplification in any NC samples, meaning adjacent sample wells are sufficiently isolated from each other. Additionally, we found no significant trend with sample location and threshold time – whether it was peripheral, central, or corner sample wells – reinforcing our previous assessment of good temperature uniformity across the device. This test was repeated in a mirrored fashion to ensure repeatability between runs despite different sample configurations. In both runs, we found a similar standard deviation between positive controls that did not correlate with position within the 96-well plate, as well as 100% accuracy of PC/NC amplification by the MINI and accompanying software.

3.4. Comparison to a commercial analyzer using two LAMP assays

To further assess the performance of MINI, we tested serial dilutions using two different LAMP assays and compared performance to a QuantStudio7 commercial qPCR device. Due to the large capacity of the MINI, we were able to run serial dilutions of both LAMP assays in triplicate within a single run [Figure 4A]. The first LAMP assay utilized KSHV plasmid samples ranging from 300,000 copies per reaction down to 96 copies per reaction, below which produces less-consistent amplification(Snodgrass et al. 2018). These samples were tested in the same plate as a second LAMP assay using GAPDH genomic DNA samples ranging from ~120,000 copies per reaction to ~100 copies per reaction. Previous testing has shown a limit of detection as low as 19 copies per reaction, however, the quantitative performance at such low copy numbers is poor compared to a different approach such as qPCR. As such, we have chosen the ~100 copies cutoff of our linear range for this analysis. Each dilution was tested in triplicate for both assays used and both assays were tested within a single run of the MINI or QuantStudio7 [Figure 4B&C]. Standard error increased at lower copy numbers of both DNA samples tested which is consistent with earlier use of these assays, and likely not due to any limitation of the MINI device as evident by the error increasing for the commercial device as well. Overall, the MINI produced comparable results to the commercially available analyzer, indicating that our unique configuration for 96 sample analysis maintains analytical performance.

Figure 4 – MINI performance compared to commercial device.

Figure 4 –

Open in a new tab

Standard curves were generated from serial dilutions analyzed in the MINI and a commercial amplification device, the QuantStudio 7. (A) Serial dilutions for both assays can be run in triplicate within a single 96-well plate. (B) GAPDH dilutions from ~120,000 copies per reaction down to the LOD of ~100 copies per reaction show similar performance between the two devices, with MINI having overall smaller error between samples. (C) KSHV plasmids ranging in concentration from 300,000 copies per reaction to 96 copies per reaction showed very similar performance and overall low error.

3.5. A high-throughput, point-of-care device for limited-resource settings

There are a number of infectious diseases that could benefit from high-throughput molecular analysis, especially in developing nations. Cervical cancer through HPV infection, vector-borne diseases like malaria, and highly transmissible diseases like tuberculosis and SARS-COV-2 all have high disease burdens in limited-resource nations that could benefit from improved access to screening technologies. Nucleic acid amplification technologies such a qPCR is commonly used for diagnosis of many infectious diseases, but is difficult to apply in limited resource settings(Obande and Singh 2020). Some success has been had with instruments like the GeneXpert by Cepheid for tuberculosis and Covid-19 testing, but this machine is still cost- and resource-demanding(Chakaya et al. 2021; Parsons et al. 2011). Many isothermal amplification technologies are under development for point-of-care testing, which have very clear benefits for use in developing nations due to device and assay simplicity, reduced cost, and accessibility(Drain et al. 2014; Moore et al. 2021; Obande and Singh 2020). However, there is a technological gap between current advances – capable of testing tens of samples per day – and the high-throughput testing – capable of hundreds of samples per day – needed to provide sufficient capacity for mass screening and point of care diagnosis.

In remote settings, decentralized approaches such as community screening programs can increase access to testing where hospitals and clinics require distant travel and high costs(Nakalembe et al. 2020). The high-capacity of the MINI device within a single run could allow multiple clinical tests to be performed within a single run, or for more detailed testing regarding a single disease. For cervical cancer screening through HPV testing, MINI offers the ability to multiplex assays and screen for multiple high-risk HPV strains. As an example, cervical swabs for eight women could be screened for 12 of the 14 HPV subtypes that are high-risk for cancer in a single run. For outbreaks of tuberculosis or malaria, and more recently COVID-19(Mozsary et al. 2021), the portable and energy-flexible MINI could offer a solution for testing large communities of people – potentially hundreds of tests in only a few hours – in non-standard settings such as schools or other public venues. Additionally, at a given reaction temperature, multiple assays can be performed within the same run due to the large number of sample wells, such as screening 48 patients for COVID-19 and tuberculosis simultaneously.

4. Conclusions

Our device MINI is the first POC accessible to device capable of analyzing 96 samples per run, and therefore hundreds of samples per day. It is well-suited for point-of-care testing as it is easily portable and simple to operate, as well as capable of maintaining functionality without additional energy input once at operating temperature, which can be realized using a variety of energy source, if necessary. We have developed a unique optical and electronic configuration to ensure complete isolation between samples, while also producing a compact POC device with limited complexity and number of electronics required. Validation was performed by testing all 96 wells simultaneously using differing arrangements of positive and negative controls. We have also shown MINI has comparable quantification performance to a commercial amplification device using two different LAMP assays for both KSHV and GAPDH. Future efforts will include design improvements to decrease costs and increase the flexibility of MINI to additional isothermal assays as well as deployment and testing in resource-limited settings. A point of care device such as MINI that can be used in decentralized healthcare settings at local clinics could provide more timely diagnoses and subsequent initiation of treatment, potentially improving health outcomes for these diseases.

Supplementary Material

supplemental data

Acknowledgements:

We would like to acknowledge our extensive list of collaborators and colleagues that provided insights during development. Additional thanks to Cornell Machine Shop manufacturing help.

DM, JB, CM, and DE received support for this work from National Institutes of Health/National Cancer Institute through a supplement to grant UH2/UH3CA202723. We also thank Testing for America (501c3), Igor Tulchinsky and the WorldQuant Foundation, Bill Ackman, Olivia Flatto and the Pershing Square Foundation, Ken Griffin and Citadel, the US National Institutes of Health (R01AI125416, R21AI129851, R01AI151059, U01DA053941).

Footnotes

Competing Interest Statement: Authors declare that they have no competing interests.

Data availability:

All data supporting the findings in this study are available within the Article and its Supplementary Information. Additional data are available from the corresponding authors upon request.

References.

  1. Biosystems A, 2003. ThermoFischer. [Google Scholar]
  2. Boehme CC, Nicol MP, Nabeta P, Michael JS, Gotuzzo E, Tahirli R, Gler MT, Blakemore R, Worodria W, Gray C, 2011. The lancet 377(9776), 1495–1505. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Butler D, Mozsary C, Meydan C, Foox J, Rosiene J, Shaiber A, Danko D, Afshinnekoo E, MacKay M, Sedlazeck FJ, 2021. Nature communications 12(1), 1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Buultjens AH, Vandelannoote K, Sharkey LK, Howden BP, Monk IR, Lee JY, Stinear TP, 2021. ACS biomaterials science & engineering 7(10), 4982–4990. [DOI] [PubMed] [Google Scholar]
  5. Chakaya J, Khan M, Ntoumi F, Aklillu E, Fatima R, Mwaba P, Kapata N, Mfinanga S, Hasnain SE, Katoto PD, 2021. International Journal of Infectious Diseases 113, S7–S12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Craw P, Balachandran W, 2012. Lab on a Chip 12(14), 2469–2486. [DOI] [PubMed] [Google Scholar]
  7. Dao Thi VL, Herbst K, Boerner K, Meurer M, Kremer LP, Kirrmaier D, Freistaedter A, Papagiannidis D, Galmozzi C, Stanifer ML, 2020. Science translational medicine 12(556), eabc7075. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Detemmerman L, Olivier S, Bours V, Boemer F, 2018. Hematology 23(3), 181–186. [DOI] [PubMed] [Google Scholar]
  9. Drain PK, Hyle EP, Noubary F, Freedberg KA, Wilson D, Bishai WR, Rodriguez W, Bassett IV, 2014. The Lancet infectious diseases 14(3), 239–249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Hagiwara M, Sasaki H, Matsuo K, Honda M, Kawase M, Nakagawa H, 2007. Journal of medical virology 79(5), 605–615. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hambalek JA, Kong JE, Brown C, Munoz HE, Horn T, Bogumil M, Quick E, Ozcan A, Di Carlo D, 2021. ACS sensors 6(9), 3242–3252. [DOI] [PubMed] [Google Scholar]
  12. Jiang M, Pan W, Arasthfer A, Fang W, Ling L, Fang H, Daneshnia F, Yu J, Liao W, Pei H, 2020. Frontiers in cellular and infection microbiology 10, 331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Kozel TR, Burnham-Marusich AR, 2017. Journal of clinical microbiology 55(8), 2313–2320. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kuhara T, Yoshikawa T, Ihira M, Watanabe D, Tamada Y, Katano H, Asano Y, Matsumoto Y, 2007. Journal of virological methods 144(1–2), 79–85. [DOI] [PubMed] [Google Scholar]
  15. Li N, Shen M, Liu J, Zhang L, Wang H, Xu Y, Cheng J, 2021a. Microsystems & nanoengineering 7(1), 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Li RJ, Mauk MG, Seok Y, Bau HH, 2021b. Analyst 146(13), 4212–4218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Liu D, Zhu Y, Li N, Lu Y, Cheng J, Xu Y, 2020a. Sensors and Actuators B: Chemical 310, 127834. [Google Scholar]
  18. Liu M, Zhao Y, Monshat H, Tang Z, Wu Z, Zhang Q, Lu M, 2020b. Biosensors and Bioelectronics 169, 112651. [DOI] [PubMed] [Google Scholar]
  19. Moore KJ, Cahill J, Aidelberg G, Aronoff R, Bektaş A, Bezdan D, Butler DJ, Chittur SV, Codyre M, Federici F, 2021. Journal of biomolecular techniques: JBT 32(3), 228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Morris U, Khamis M, Aydin-Schmidt B, Abass AK, Msellem MI, Nassor MH, González IJ, Mårtensson A, Ali AS, Björkman A, 2015. Malaria journal 14(1), 1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mozsary C, McCloskey D, Babler KM, Boza J, Butler D, Currall B, Williams S, Wiley A, Afshin EE, Grills GS, 2021. Journal of Biomolecular Techniques: JBT 32(3), 221. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Nakalembe M, Makanga P, Kambugu A, Laker-Oketta M, Huchko MJ, Martin J, 2020. Cancer medicine 9(22), 8701–8712. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Nakisige C, Trawin J, Mitchell-Foster S, Payne BA, Rawat A, Mithani N, Amuge C, Pedersen H, Orem J, Smith L, 2020. BMC public health 20(1), 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Nhu NTQ, Heemskerk D, Thu DDA, Chau TTH, Mai NTH, Nghia HDT, Loc PP, Ha DTM, Merson L, Thinh TTV, 2014. Journal of clinical microbiology 52(1), 226–233. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nicol MP, Workman L, Isaacs W, Munro J, Black F, Eley B, Boehme CC, Zemanay W, Zar HJ, 2011. The Lancet infectious diseases 11(11), 819–824. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Obande GA, Singh KKB, 2020. Infection and drug resistance 13, 455. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Oeschger TM, McCloskey DS, Buchmann RM, Choubal AM, Boza JM, Mehta S, Erickson D, 2021. Accounts of Chemical Research 54(19), 3656–3666. [DOI] [PubMed] [Google Scholar]
  28. Panpradist N, Kline EC, Atkinson RG, Roller M, Wang Q, Hull IT, Kotnik JH, Oreskovic AK, Bennett C, Leon D, 2021. Science advances 7(51), eabj1281. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Parsons LM, Somoskövi Á, Gutierrez C, Lee E, Paramasivan C, Abimiku A.l., Spector S, Roscigno G, Nkengasong J, 2011. Clinical microbiology reviews 24(2), 314–350. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Ponce C, Kaczorowski F, Perpoint T, Miailhes P, Sigal A, Javouhey E, Gillet Y, Jacquin L, Douplat M, Tazarourte K, 2017. Parasite 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Rohatensky MG, Livingstone DM, Mintchev P, Barnes HK, Nakoneshny SC, Demetrick DJ, Dort JC, van Marle G, 2018. BMC cancer 18(1), 1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Snodgrass R, Gardner A, Semeere A, Kopparthy VL, Duru J, Maurer T, Martin J, Cesarman E, Erickson D, 2018. Nature biomedical engineering 2(9), 657–665. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F, 2021. CA: a cancer journal for clinicians 71(3), 209–249. [DOI] [PubMed] [Google Scholar]
  34. Tekola-Ayele F, Rotimi CN, 2015. Public health genomics 18(4), 242–247. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang Y, Li K, Xu G, Chen C, Song G, Dong Z, Lin L, Wang Y, Xu Z, Yu M, 2021. Research 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wei Z, Wang X, Feng H, Ji F, Bai D, Dong X, Huang W, 2022. Critical Reviews in Biotechnology, 1–18. [DOI] [PubMed] [Google Scholar]
  37. Wu DY, Ugozzoli L, Pal BK, Wallace RB, 1989. Proceedings of the National Academy of Sciences 86(8), 2757–2760. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Yager P, Domingo GJ, Gerdes J, 2008. Annu. Rev. Biomed. Eng. 10, 107–144. [DOI] [PubMed] [Google Scholar]
  39. Yin K, Pandian V, Kadimisetty K, Ruiz C, Cooper K, You J, Liu C, 2019. Theranostics 9(9), 2637. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Yu L, Wu S, Hao X, Dong X, Mao L, Pelechano V, Chen W-H, Yin X, 2020a. Clinical chemistry 66(7), 975–977. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yu Z, Lyu W, Yu M, Wang Q, Qu H, Ismagilov RF, Han X, Lai D, Shen F, 2020b. Biosensors and Bioelectronics 155, 112107. [DOI] [PubMed] [Google Scholar]
  42. Zhang H, Xu Y, Fohlerova Z, Chang H, Iliescu C, Neuzil P, 2019. TrAC Trends in Analytical Chemistry 113, 44–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Zhang M, Liu J, Shen Z, Liu Y, Song Y, Liang Y, Li Z, Nie L, Fang Y, Zhao Y, 2021. BMC microbiology 21(1), 1–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Zhu X, Wang X, Han L, Chen T, Wang L, Li H, Li S, He L, Fu X, Chen S, 2020. Biosensors and Bioelectronics 166, 112437. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental data
NIHMS1970248-supplement-supplemental_data.docx (16.2KB, docx)

Data Availability Statement

All data supporting the findings in this study are available within the Article and its Supplementary Information. Additional data are available from the corresponding authors upon request.

RESOURCES