Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 8.
Published in final edited form as: Anal Chim Acta. 2024 Feb 17;1297:342371. doi: 10.1016/j.aca.2024.342371

Improving Bacteria Identification from Digital Melt Assay via Oligonucleotide-Based Temperature Calibration

Amelia Traylor a,‡, Pei-Wei Lee a,‡, Kuangwen Hsieh a, Tza-Huei Wang a,b
PMCID: PMC11082877  NIHMSID: NIHMS1970192  PMID: 38438240

Abstract

Background:

Bacterial infections, especially polymicrobial infections, remain a threat to global health and require advances in diagnostic technologies for timely and accurate identification of all causative species. Digital melt – microfluidic chip-based digital PCR combined with high resolution melt (HRM) – is an emerging method for identification and quantification of polymicrobial bacterial infections. Despite advances in recent years, existing digital melt instrumentation often delivers nonuniform temperatures across digital chips, resulting in nonuniform digital melt curves for individual bacterial species. This nonuniformity can lead to inaccurate species identification and reduce the capacity for differentiating bacterial species with similar digital melt curves.

Results:

We introduce herein a new temperature calibration method for digital melt by incorporating an unamplified, synthetic DNA fragment with a known melting temperature as a calibrator. When added at a tuned concentration to an established digital melt assay amplifying the commonly targeted 16S V1 – V6 region, this calibrator produced visible low temperature calibrator melt curves across-chip along with the target bacterial melt curves. This enables alignment of the bacterial melt curves and correction of heating-induced nonuniformities. Using this calibration method, we were able to improve the uniformity of digital melt curves from three causative species of bacteria. Additionally, we assessed calibration’s effects on identification accuracy by performing machine learning identification of three polymicrobial mixtures comprised of two bacteria with similar digital melt curves in different ratios. Calibration greatly improved mixture composition prediction.

Significance:

To the best of our knowledge, this work represents the first DNA calibrator-supplemented assay and calibration method for nanoarray digital melt. Our results suggest that this calibration method can be flexibly used to improve identification accuracy and reduce melt curve variabilities across a variety of pathogens and assays. Therefore, this calibration method has the potential to elevate the diagnostic capabilities of digital melt toward polymicrobial bacterial infections and other infectious diseases.

Keywords: Digital melt, Calibration, Nanoarray microfluidics, Nucleic acid detection

1. Introduction

Bacterial infections are the underlying cause for many prevalent and often life-threatening illnesses [1,2]. In particular, bacterial infections that involve multiple causative species – or polymicrobial infections – have been observed in bloodstream infections [3,4] urinary tract infections [5,6] and wound infections [7,8] and are associated with higher infection severity and worse patient outcomes [9–13]. Timely and accurate identification of all causative species for initiating appropriate treatment is essential to patient outcomes [14–16]. However, current diagnostic methods rely mainly on traditional bacterial culture – a time-consuming method that can take up to 5 days [17,18]. Polymerase chain reaction (PCR) combined with high resolution melt (HRM) has become a commonly researched method for timely identification of bacteria [19–22]. Still, these bulk-based melt assays cannot analyze highly polymicrobial samples, due to the formation of composite melt curves that are difficult to resolve.

To address this concern, nanoarray-based digital melt platforms have been proposed, in which a sample is digitized via a microfluidic chip into wells containing single copies of bacterial DNA [23–25]. While this method allows broad bacterial identification in polymicrobial samples, correct classification of bacteria with similar melt curves can be difficult due to small melt curve shifts along the temperature axis caused by heating variations across the digital chip [26,27]. These variations result from imperfections in the heating system or poor contact between the heating block and the chip. Therefore, there is a need for an internal temperature calibration method for digital melt curves to correct for these heating variations. Oligonucleotides have been used as temperature calibrators in tandem with bulk melt assays and microtiter-plate-based quasi-digital melt assays [28–32]. They have also been employed for system temperature calibration in microfluidic chip platforms [23,26]. To date, however, oligonucleotide-based temperature calibrators have yet to be incorporated into microfluidic digital melt assays for melt curve calibration.

Herein, we demonstrate that supplementation of nanoarray digital melt assays with an internal, unamplified DNA calibrator facilitates digital melt curve alignment and improves bacterial identification accuracy by reducing instrumentation-induced heating nonuniformities (Fig. 1). When calibrator is spiked into the sample at a high concentration, each nanowell of the high density nanoarray contains calibrator, producing a melt curve of known TM. This known calibrator curve can be aligned and used to shift nearby bacterial melt curves, therefore reducing the effects of a spatial temperature gradient on melt curve alignment. By improving the resolvable distance between similar melt curves, this calibration method could expand nanoarray digital melt’s capability for accurate identification of bacteria, addressing a concerning gap in current diagnostics for rapid identification of severe polymicrobial infections.

Fig. 1.

Fig. 1.

Open in a new tab

A sample containing bacterial DNA is digitized onto a high density nanoarray, where digital melt is performed. The resulting melt curves show unwanted variation in position on the temperature axis due to imperfections in the heating system that result in a cross-chip temperature gradient. Introducing a synthetic DNA calibrator with known melt temperature in high concentration to the nanoarray could reduce the effects of heating variation on bacterial species-specific melt curves, resulting in better curve alignment and, therefore, improved identification ability.

2. Materials and Methods

2.1. Bacterial Strains, DNA Extraction, and Target Preparation

Three bacterial strains were sourced from the American Type Culture Collection (ATCC; Manassas, VA). These strains were Acinetobacter baumannii (A. baumannii) ATCC 19606, Klebsiella pneumoniae (K. pneumoniae) BAA-1705, and Morganella morganii (M. morganii) ATCC 8076H. Bacterial DNA was extracted from these bacteria isolates using QuickExtract Bacterial DNA Extraction Kit (Epicentre, Madison, WI) according to the manufacturer’s protocol. The extracted bacterial DNA was stored at −20 °C and diluted before loading to a 5040-well nanoarray. To achieve full digitization, the mean occupancy of DNA per nanowell (λ) should be less than 0.1 [24]. After developing a standard curve for each bacterial DNA extraction, dilutions of fully digitizable DNA were stored at 4 °C for the rest of the experimentation period to prevent DNA fragmentation during freeze-thaw cycles, which could result in more positive wells than genomic copies.

2.2. Development of Digital Melt Assays with dsDNA Calibrator

Three primer sets targeting regions around the 16S gene and ITS region, which are commonly used for bacterial identification [33–37], were modified to include calibrator in this work. The V1 – V6 primer set targets the V1 – V6 hypervariable portions of 16S gene. The tRNAala primer covers the ITS sequence between 16S gene and the tRNAala gene. Finally, the ITS primer set targets the ITS region between the 16S and 23S genes.

For internal calibration of our digital melt platform, we chose a DNA calibrator with low melting temperature (~65 °C) that has been previously validated for melt curve calibration in bulk [28]. This calibrator sequence includes a three-carbon alkyl group at their 3’ end to prevent calibrator extension during PCR. In previous literature, two calibrators, one with low melting temperature and one with high melting temperature, have been used for temperature calibration [28]. However, the established high temperature calibrator sequence is not suitable for bacterial identification applications, since bacterial product melt curves overlap with its melting temperature of around 92 °C. Therefore, we proceeded with calibration using only a low temperature calibrator.

2.3. Bulk-based Melt Curve Generation

For bulk-based melt, the PCR mixture was composed of 1 μL diluted extracted bacterial DNA, 1× Gold Buffer (Thermo Fisher Scientific, Waltham, MA), 3.5 mM MgCl2 (Thermo Fisher Scientific, Waltham, MA), 1× Evagreen (Biotium, Freemont, CA), 1 mg mL−1 BSA (New England Biolabs, Ipswich, MA), 0.01% Tween 20 (Sigma-Aldrich, St. Louis, MO), 0.3 μM forward and reverse primers targeting V1-V6, ITS or tRNAala (Integrated DNA Technologies, Coralville, IA), 200 μM of each deoxynucleotide triphosphate (Thermo Fisher Scientific, Waltham, MA), 0.05 U μL−1 of Amplitaq Gold LD (Thermo Fisher Scientific, Waltham, MA), and Ultra Pure PCR water (Quality Biological Inc., Gaithersburg, MD) to a final reaction volume of 10 μL. Bulk-based PCR and melt were performed for this mix on the CFX Real-Time PCR detection system (Bio-Rad, Hercules, CA). All PCR assays started with a 95 °C hot start for 10 min followed by 60 amplification cycles, and a final extension at 72 °C for 7 min. The amplification cycles for mixes containing 16S V1 – V6 primers were each 95 °C for 15 s, 65 °C for 15 s, and 72 °C for 60 s. Meanwhile, the amplification cycles for mixes containing ITS or tRNAala primers were each 95 °C for 15 s, 60 °C for 15 s, and 72 °C for 60 s. Post-PCR DNA melting was then completed on all mixes from 45°C to 95°C with 0.1°C sec−1 ramping rate to obtain fluorescence versus time curves. The negative derivative of these fluorescence curves was then plotted in MATLAB.

2.4. Digital Melt Curve Generation

Our digital melt platform for broad bacterial ID centers around a previously described microfluidic nanoarray, on which PCR mixtures are vacuum loaded and digitized using an oil mixture [12–14]. The PCR mixture for digital PCR was composed of 2.5 μL diluted extracted bacterial DNA, 1× Gold Buffer (Thermo Fisher Scientific, Waltham, MA), 3.5 mM MgCl2 (Thermo Fisher Scientific, Waltham, MA), 1× Evagreen (Biotium, Freemont, CA), 1 mg mL−1 BSA (New England Biolabs, Ipswich, MA), 0.01% Tween 20 (Sigma-Aldrich, St. Louis, MO), 0.3 μM forward and reverse primers targeting V1-V6, ITS or tRNAala (Integrated DNA Technologies, Coralville, IA), 200 μM of each deoxynucleotide triphosphate (Thermo Fisher Scientific, Waltham, MA), 0.05 U μL−1 of Amplitaq Gold LD (Thermo Fisher Scientific, Waltham, MA), and Ultra Pure PCR water (Quality Biological Inc., Gaithersburg, MD) to a final reaction volume of 25 μL. The mixture-containing nanoarray is then placed on a thermal platform for dPCR thermocycling and digital melt temperature ramping. The thermal cycling conditions for digital PCR are identical to the described thermal cycling for bulk PCR. Post-PCR DNA melting was then completed from 45 °C to 99.9 °C with a 0.1 °C sec−1 ramping rate. During this melting, fluorescence images are captured every two seconds and are then post-processed to produce temperature-lapsed digital melt curves for each nanowell.

2.5. Calibration of Digital Melt Curves

Once digital melt curves are obtained, wells containing bacterial product are identified by selecting curves that contain a peak above a threshold height within the expected temperature range. All bacteria-positive digital melt curves are then normalized by the area under each curve. Before calibration, all bacteria-negative wells that contain a calibrator peak are shifted along the temperature axis so that calibrator curves are aligned at a fixed point, defined by the bulk melting temperature of the calibrator. Then, the melt curve for each bacteria-positive well is translated by a shift factor calculated by taking the average alignment shift for all bacteria-negative wells directly surrounding the bacteria-positive well.

While heating imperfections are a major cause of melt curve variations, additional factors including variations in DNA quality and local differences in reagent concentrations between wells also can result in small differences in melt peak [40]. By taking the average of surrounding negative well alignment shift values rather than simply aligning the calibrator peak in bacteria-positive wells, we reduce the impact of variations from these additional factors on our calibration shift.

2.6. Machine-Learning Assisted Algorithm for Bacterial ID

To test bacterial identification accuracy, digital melt curves for K. pneumoniae, M. morganii, and a mixture of the two species were generated simultaneously on separate modules of a single digital chip. Each module also contained calibrator. All bacterial melt curves were normalized by the area under the curve, saved before and after calibration, and then cropped to 75 °C – 99 °C to remove any residual calibrator effects on the melt curve. The processed melt curves from single-bacteria modules were then used to train a one-versus-one SVM (ovo-SVM) algorithm [41]. This algorithm converts these test curves into feature vectors and calculates a margin that represents the maximum separation of feature vectors with different classifications using a linear kernel function and the least squares method. The accuracy of the trained algorithm was then tested with the polymicrobial module melt curves. Our algorithm employs a binary assessment approach, wherein each feature vector derived from the bacterial digital melt curve is compared against the SVM margin calculated from the training digital melt curves. The outcome of this comparison yields scores for each digital melt curve, quantifying the frequency of classification as each respective species. Ultimately, the highest matching score from the ovo-SVM outcomes determines the identification of the bacterial species.

3. Results and Discussion

3.1. Initial Evaluation of Calibrator via Bulk-based Melt Assays

Effective temperature calibration for digital melt entails using a calibrator that is compatible with digital melt assays at a concentration that can produce visible calibrator melt curves across the entire digital chip without inhibiting target melt curves. When these requirements are met, the calibrator melt temperatures serve as a map of cross-chip spatial temperature difference and allow for the correction of any temperature gradient. We began with three of our commonly used digital melt assays with distinct primer sets targeting different gene regions – 16S V1 - V6, tRNAala, and ITS [24,38,39] – and used A. baumannii as a target causative pathogen, since it is often found in intensive care units, poses a high mortality risk, and has developed antibiotic resistance [42]. We then identified a previously reported synthetic double-stranded DNA fragment [28] to use as the calibrator. This calibrator lacks complementary sequences to our primers and target sequences, and therefore should not be amplified or affect target melt curves. Based on our initial benchtop evaluation using a target concentration of 1000 target copies μL−1 (approximating 1 copy per 1nL digital well) and various calibrator concentrations, we determined that 0.2 μM calibrator for the 16S V1 - V6 assay, 0.1 μM calibrator for the tRNAala assay, and 0.05 μM calibrator for the ITS assay produced appropriate calibrator melt curves and target melt curves (Fig. S1). At these calibrator concentrations, we observed some delay in cycles of quantification (Cq) for all three assays, signaling PCR inhibition (Fig. S2). However, since small amplification delays do not impact our quantification ability using digital melt, we still proceeded to implement all three assays in digital chips.

3.2. Implementation and Evaluation of Calibrator-Supplemented Digital Melt Assays

For implementation of digital melt calibration, we began by performing and comparing calibrator-supplemented and calibrator-free assays in all three assays using our standard digital chip and digital melt platform [24]. In addition to bright positive wells containing bacterial DNA, all wells in the calibrator-supplemented 16S V1 – V6 assay show an elevated level of fluorescence at low temperatures due to the presence of calibrator DNA. As the temperature increases past 65 °C, fluorescence in wells containing only calibrator disappears, indicating calibrator melting. As the temperature further increases past 92 °C, fluorescence in bacteria-positive wells also disappears, indicating target melting (Fig. 2a, top). In the calibrator-free 16S V1 – V6 assay, strongly fluorescent bacteria-positive wells were observed at low temperatures with minimal signal from other wells. The only large drop in fluorescence during this melt occurs around 92 °C, when fluorescence in the bacteria-positive wells disappears, indicating target melting (Fig. 2a, bottom). Both the tRNAala and ITS assays show similar differences between the calibrator-supplemented and calibrator-free modules (Fig. S3).

Fig. 2.

Fig. 2.

Open in a new tab

(a) Images of a nanoarray portion during V1-V6 digital melt for diluted A. baumannii DNA, with or without added DNA calibrator, show that we can distinguish bright positive wells containing A. baumannii from background for both conditions. At low temperatures, wells in the calibrator-containing module also fluoresce, suggesting calibrator is present in all wells. (b) Digital melt curves for three different assays using 16S V1 – V6, tRNAala, and ITS primers with and without added calibrator show that addition of calibrator does not affect positive well count or bacterial melt curve variability for 16s V1 – V6 or tRNAala primers, while calibrator addition reduces positive well count when using ITS primers.

The digital melt curves extracted from the time lapse fluorescence images of the calibrator-supplemented 16S V1 – V6 assay show clear calibrator curves with melting temperatures (TM) of ~65 °C, which are distinct from A. baumannii curves with TM of ~92 °C (Fig. 2b, top left). In contrast, we observed only A. baumannii curves from the calibrator-free 16S V1 – V6 assay (Fig. 2b, bottom left). Importantly, both the number of A. baumannii melt curves and the TM standard variations are comparable between the two assays, suggesting the bulk-tuned concentration of calibrator is appropriately low enough to minimize interaction with the bacterial melt curves. Also, in the calibrator-supplemented assay, calibrator curves are not visible in wells with A. baumannii present. Since the bacterial DNA provides a high background fluorescence in these wells at the calibrator melting temperature, the signal drop from calibrator melting does not provide a strong melt curve. Adding additional calibrator could prevent this effect but would result in an undesirable higher impact on bacterial melt curves.

We subsequently performed calibrator-supplemented and calibrator-free tRNAala and ITS digital melt assays. For the tRNAala digital melt assays, we observed both calibrator curves and A. baumannii curves from the calibrator-supplemented assay (Fig. 2b, top middle) but only A. baumannii curves from the calibrator-free assay (Fig. 2b, bottom middle). The calibrator curves have lower peak heights than those of the 16S V1 – V6 assay, consistent with the lower calibrator concentration used in the tRNAala assay. For the ITS digital melt assays, we observed both calibrator curves and double-peak curves associated with A. baumannii from the calibrator-supplemented assay (Fig. 2b, top right) but only A. baumannii double-peak curves from the calibrator-free assay (Fig. 2b, bottom right). The calibrator curves were again lower than those of the 16S V1 – V6 assay due to the lower calibrator concentration used in the ITS assay. Unlike the other two calibrator-supplemented assays, small calibrator curves are visible in some wells containing A. baumannii. Also, the number of A. baumannii melt curves from the calibrator-supplemented assay was notably 62.6% lower than the calibrator-free assay, suggesting inhibition to the ITS assay with calibrator addition. Such inhibition presumably arose due to nonspecific interactions between the degenerate bases of the ITS primers and the calibrator, which led to poor PCR efficiency. Based on these results, we elected to proceed with our study using the calibrator-supplemented 16S V1 – V6 digital melt assay, which provided the strongest calibrator signal among the three assays.

3.3. Temperature Calibration Improves Uniformity in Bacterial Digital Melt Curves

We then developed a method for aligning bacterial melt curves on the temperature axis based on their surrounding calibrator curves (Fig. 3a). In our method, we first determine TM for calibrator melt curves and bacterial melt curves from all wells in the digital chip (Step 1). When using our tuned calibrator concentration to minimize calibrator impact on bacterial melt curves, we do not see a visible calibrator peak in bacteria-containing wells. Therefore, we cannot directly align bacterial melt curves with a calibrator melt curve from the same well. Instead, we use the calibrator melt curves from wells surrounding a bacterial melt curve as a proxy for the calibrator melt curve at that location. For these neighboring calibrator curves, we determine their temperature differences (denoted as δ) from the reference TM of the calibrator, as measured in the bulk-based melt assay, and calculate the average temperature difference (i.e., mean δ) of these bacteria-surrounding calibrator curves (Step 2). If a bacteria-positive well has a neighbor that also contains bacterial product, we do not include that neighbor in calibration calculations. We subsequently shift each bacterial melt curve by the mean δ from its surrounding calibrator curves (Step 3). In doing so, we can align bacterial melt curves and reduce any variations across the digital chip due to temperature nonuniformity of the instrumentation (Step 4).

Fig. 3.

Fig. 3.

Open in a new tab

(a) Calibration is performed by translating bacterial melt curves on the temperature axis by the average reference shift for surrounding calibrator melt curves multiplied by a shift factor that corrects for distribution differences between the calibrator and the product. (b) Melt curves and TM distributions from digital melt of A. baumannii (n = 291), K. pneumoniae (n = 188), and M. morganii (n = 355) are shown before and after temperature calibration. Temperature calibration reduced melt curve variability for all three bacteria.

We subsequently demonstrated the advantage of temperature-calibrated digital melt for multiple bacteria species. Here, in addition to A. baumannii, we also targeted two other pathogenic bacteria in K. pneumoniae and M. morganii. For these three targets, we performed calibrator-supplemented digital melt and compared their bacterial melt curve before and after temperature calibration. For all three targets, their bacterial melt curves were more aligned upon temperature calibration (Fig. 3b). The full digital melt curves produced during these experiments can be found in Fig. S4a. Heatmaps of the bacterial melting temperatures (TMs), which represent the spatial temperature distribution of each curve’s maximum peak across-chip, showed a clear temperature gradient across the pre-calibration modules. This pattern verifies that cross-platform heating variations occur in the system. Calibration reduces this cross-chip TM gradient, suggesting that these alignment improvements are correlated with a reduction in heating variability. Finally, to assess TM variation, we looked at two measures: the standard deviation of the TMs and the percentage of TMs clustered around the median. For this second metric, we included any curves with TM within the smallest resolvable temperature step (0.2 °C) of the median. Calibration of A. baumannii reduces standard deviation from 0.345 to 0.168, while increasing melt peak clustering from 32% to 95.9%. Calibration of K. pneumoniae reduces standard deviation from 0.212 to 0.156, while increasing melt peak clustering from 88.3% to 92.6%. Finally, calibration of M. morganii reduces standard deviation from 0.281 to 0.213, while increasing melt peak clustering from 58.9% to 85.1%. These results show that temperature calibration for digital melt is applicable for multiple bacteria of interest.

3.4. Temperature Calibration Enhances Identification Accuracy of Polymicrobial Mixture

After demonstrating that calibration results in improved melt peak alignment, we developed an experiment to test calibration’s ability to improve machine-learning (ML) based identification of species with similar melt curves. M. morganii and K. pneumoniae produce similar melt curves using the 16S V1 – V6 digital melt assay, with melt temperatures separated by only about 1 °C. These species are therefore difficult to identify in a polymicrobial mixture due to cross-chip melt curve variations. By reducing these variations, calibration makes these species easier for the ML algorithm to separate.

To test this hypothesis, we ran three-module calibrator-supplemented digital melt experiments, with one module containing M. morganii, one module containing K. pneumoniae, and one module containing a mixture of the two species (Fig. 4a). We ran three of these experiments, each with a different two-species mixture expected ratio (1:1, 4:1, and 1:6). The full digital melt curves produced during these experiments can be found in Fig. S4b. For each of the single-bacteria modules, we used all the resulting bacterial melt curves as the training set for an OVO-SVM model. The bacteria-positive well count of these single-bacteria modules were used along with the volume of each target added to finalize the expected mixture ratio. We then identified the polymicrobial module bacterial melt curves using the trained model. Training and testing the model on uncalibrated melt curves led to poor identification of the polymicrobial melt curves, while calibration resulted in improved identification accuracy (Fig. 4b). Of the uncalibrated polymicrobial module melt curves, the algorithm classifies the 1:1 mixture as 1:23.4 K. pneumoniae to M. morganii, the 4:1 mixture as 1:33.5 K. pneumoniae to M. morganii, and the 1:6 mixture as 1:126.5 K. pneumoniae to M. morganii. This poor classification is likely due to module-to-module and well-to-well variations in heating that result in misidentification. Calibration results in superior composition prediction, with the 1:1 mixture identified as 1:2.1 K. pneumoniae to M. morganii, the 4:1 mixture identified as 2.58:1 K. pneumoniae to M. morganii, and the 1:6 mixture identified as 1:7.23 K. pneumoniae to M. morganii.

Fig. 4.

Fig. 4.

Open in a new tab

(a) We obtained bacterial melt curves for three independent three-module digital chips. Each chip had one module containing K. pneumoniae, one module containing M. morganii, and one module containing a polymicrobial mixture of various ratios (1:1, 4:1, or 1:6). The single-bacteria module melt curves were used to train an OVO-SVM identification algorithm, which we tested on the polymicrobial module curves. (b) Test melt curves from the polymicrobial mixtures are grouped by species identification from an OVO-SVM algorithm. Calibration results in more accurate composition identification.

4. Conclusions

To improve bacterial identification accuracy from digital melt, we introduce a temperature calibration method for digital melt platforms that uses a DNA calibrator with known TM to correct instrumentation induced heating variations and reduce bacterial melt curve variability. We verify that calibrator addition is compatible with three established digital melt assays targeting different gene regions. In particular, we determine the digital melt assay targeting the 16S V1 – V6 region is highly amenable to calibrator supplementation, with no observed negative impact on bacterial melt curves due to calibrator addition. After tuning an appropriate calibrator concentration to provide visible calibration curves across the entire digital chip without negatively impacting the digital melt assay, we show that calibration improves melt curve alignment from the 16S V1 – V6 region for three causative pathogens. Additionally, we demonstrate that composition identification of polymicrobial mixtures of two bacteria with similar melt profiles is improved post-calibration. These results suggest that this calibration method can be flexibly used to improve identification accuracy and reduce melt curve variabilities across a variety of pathogens and assays.

To the best of our knowledge, our work represents the first DNA calibrator-supplemented digital melt assay. As such, we foresee several opportunities for improvements and follow-up studies. For example, to further validate our calibration method, we would test calibration and subsequent identification using a larger sample of clinically relevant bacteria. Additionally, increasing the number of bacterial melt curves used as training data could improve identification accuracy. We also would like to test other calibrator sequences of different TM in combination with the current sequence to improve calibration performance by accounting for cross-platform variations in heating rate. Finally, we believe that our calibration method can be implemented in other digital melt platforms, including droplet platforms and commercially available array platforms. Based on our current results and the potential for improvement, we envision oligonucleotide-based temperature calibration being widely implemented in digital melt platforms to improve their capabilities as rapid, accurate diagnostics for polymicrobial infections.

Supplementary Material

1

Table 1.

Calibrator and Primer Sequences.

Sequence
Calibrator 5’-TTAATTATAAAGGTATTTATAATATTGAATTATACATATCTAATATAATC-C3–3’[28]
V1 – V6 Primers Forward
Reverse		 5’-GYGGCGNACGGGTGAGTAA-3’[24]
5’-AGCTGACGACANCCATGCA-3’[24]
tRNAala Primers Forward
Reverse		 5’-AAGTCGTAACAAGGTARCCG-3’[38]
5’-ACCYCCTGCKTGCAARGCA-3’[38]
ITS Primers Forward
Reverse		 5’-CGGTGAATACGTTCCCGGIIIIIGTAC −3’[39]
5’-CGTCCTTCDTCGVCTBIIIIIGCCARG-3’[39]
Open in a new tab

Note: Y, N, D, V, B, R, and K refer to degenerate bases where Y = C, T; N = A, C, G, T; D = A, G, T; V = A, C, G; B = G, C, T; R = A, G; K = G, T; I refers to inosine

Highlights.

  • Digital melt is an emerging method for identifying tough polymicrobial infections

  • Cross-platform temperature variations reduce digital melt identification accuracy

  • DNA temperature calibrator enables alignment and improves uniformity of melt curves

  • Calibrated digital melt shows enhanced polymicrobial sample identification accuracy

Acknowledgements

This research is financially supported by the National Institutes of Health [R01AI137272, R01CA260628, and R33CA272321] and the National Cancer Institute [T32CA153952].

Footnotes

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • [1].Ikuta KS, Swetschinski LR, Aguilar G. Robles, Sharara F, Mestrovic T, Gray AP, Weaver N. Davis, Wool EE, Han C, Hayoon A. Gershberg, Aali A, Abate SM, Abbasi-Kangevari M, Abbasi-Kangevari Z, Abd-Elsalam S, Abebe G, Abedi A, Abhari AP, Abidi H, Aboagye RG, Absalan A, Abubaker Ali H, Acuna JM, Adane TD, Addo IY, Adegboye OA, Adnan M, Adnani QES, Afzal MS, Afzal S, Aghdam ZB, Ahinkorah BO, Ahmad A, Ahmad AR, Ahmad R, Ahmad S, Ahmad S, Ahmadi S, Ahmed A, Ahmed H, Ahmed JQ, Rashid T. Ahmed, Ajami M, Aji B, Akbarzadeh-Khiavi M, Akunna CJ, Al Hamad H, Alahdab F, Al-Aly Z, Aldeyab MA, Aleman AV, Alhalaiqa FAN, Alhassan RK, Ali BA, Ali L, Ali SS, Alimohamadi Y, Alipour V, Alizadeh A, Aljunid SM, Allel K, Almustanyir S, Ameyaw EK, Amit AML, Anandavelane N, Ancuceanu R, Andrei CL, Andrei T, Anggraini D, Ansar A, Anyasodor AE, Arabloo J, Aravkin AY, Areda D, Aripov T, Artamonov AA, Arulappan J, Aruleba RT, Asaduzzaman M, Ashraf T, Athari SS, Atlaw D, Attia S, Ausloos M, Awoke T, Quintanilla B.P. Ayala, Ayana TM, Azadnajafabad S, Jafari A. Azari, B DB, Badar M, Badiye AD, Baghcheghi N, Bagherieh S, Baig AA, Banerjee I, Barac A, Bardhan M, Barone-Adesi F, Barqawi HJ, Barrow A, Baskaran P, Basu S, Batiha A-MM, Bedi N, Belete MA, Belgaumi UI, Bender RG, Bhandari B, Bhandari D, Bhardwaj P, Bhaskar S, Bhattacharyya K, Bhattarai S, Bitaraf S, Buonsenso D, Butt ZA, dos Santos F.L. Caetano, Cai J, Calina D, Camargos P, Cámera LA, Cárdenas R, Cevik M, Chadwick J, Charan J, Chaurasia A, Ching PR, Choudhari SG, Chowdhury EK, Chowdhury FR, Chu D-T, Chukwu IS, Dadras O, Dagnaw FT, Dai X, Das S, Dastiridou A, Debela SA, Demisse FW, Demissie S, Dereje D, Derese M, Desai HD, Dessalegn FN, Dessalegni SAA, Desye B, Dhaduk K, Dhimal M, Dhingra S, Diao N, Diaz D, Djalalinia S, Dodangeh M, Dongarwar D, Dora BT, Dorostkar F, Dsouza HL, Dubljanin E, Dunachie SJ, Durojaiye OC, Edinur HA, Ejigu HB, Ekholuenetale M, Ekundayo TC, El-Abid H, Elhadi M, Elmonem MA, Emami A, Bain L. Engelbert, Enyew DB, Erkhembayar R, Eshrati B, Etaee F, Fagbamigbe AF, Falahi S, Fallahzadeh A, Faraon EJA, Fatehizadeh A, Fekadu G, Fernandes JC, Ferrari A, Fetensa G, Filip I, Fischer F, Foroutan M, Gaal PA, Gadanya MA, Gaidhane AM, Ganesan B, Gebrehiwot M, Ghanbari R, Ghasemi Nour M, Ghashghaee A, Gholamrezanezhad A, Gholizadeh A, Golechha M, Goleij P, Golinelli D, Goodridge A, Gunawardane DA, Guo Y, Gupta RD, Gupta S, Gupta VB, Gupta VK, Guta A, Habibzadeh P, Haddadi Avval A, Halwani R, Hanif A, Hannan Md.A., Harapan H, Hassan S, Hassankhani H, Hayat K, Heibati B, Heidari G, Heidari M, Heidari-Soureshjani R, Herteliu C, Heyi DZ, Hezam K, Hoogar P, Horita N, Hossain MM, Hosseinzadeh M, Hostiuc M, Hostiuc S, Hoveidamanesh S, Huang J, Hussain S, Hussein NR, Ibitoye SE, Ilesanmi OS, Ilic IM, Ilic MD, Imam MT, Immurana M, Inbaraj LR, Iradukunda A, Ismail NE, Iwu CCD, Iwu CJ, J LM, Jakovljevic M, Jamshidi E, Javaheri T, Javanmardi F, Javidnia J, Jayapal SK, Jayarajah U, Jebai R, Jha RP, Joo T, Joseph N, Joukar F, Jozwiak JJ, Kacimi SEO, Kadashetti V, Kalankesh LR, Kalhor R, Kamal VK, Kandel H, Kapoor N, Karkhah S, assa BG, Kassebaum NJ, Katoto PD, Keykhaei M, Khajuria H, Khan A, Khan IA, Khan M, Khan MN, Khan MA, Khatatbeh MM, Khater MM, Kashani H.R. Khayat, Khubchandani J, Kim H, Kim MS, Kimokoti RW, Kissoon N, Kochhar S, Kompani F, Kosen S, Koul PA, Laxminarayana S.L. Koulmane, Krapp Lopez F, Krishan K, Krishnamoorthy V, Kulkarni V, Kumar N, Kurmi OP, Kuttikkattu A, Kyu HH, Lal DK, Lám J, Landires I, Lasrado S, Lee S, Lenzi J, Lewycka S, Li S, Lim SS, Liu W, Lodha R, Loftus MJ, Lohiya A, Lorenzovici L, Lotfi M, Mahmoodpoor A, Mahmoud MA, Mahmoudi R, Majeed A, Majidpoor J, Makki A, Mamo GA, Manla Y, Martorell M, Matei CN, McManigal B, Mehrabi Nasab E, Mehrotra R, Melese A, Mendoza-Cano O, Menezes RG, Mentis A-FA, Micha G, Michalek IM, Nogueira de Sá A.C. Micheletti Gomide, Kostova N. Milevska, Mir SA, Mirghafourvand M, Mirmoeeni S, Mirrakhimov EM, Mirza-Aghazadeh-Attari M, Misganaw AS, Misganaw A, Misra S, Mohammadi E, Mohammadi M, Mohammadian-Hafshejani A, Mohammed S, Mohan S, Mohseni M, Mokdad AH, Momtazmanesh S, Monasta L, Moore CE, Moradi M, Moradi Sarabi M, Morrison SD, Motaghinejad M, Mousavi Isfahani H, Mousavi Khaneghah A, Mousavi-Aghdas SA, Mubarik S, Mulita F, Mulu GBB, Munro SB, Muthupandian S, Nair TS, Naqvi AA, Narang H, Natto ZS, Naveed M, Nayak BP, Naz S, Negoi I, Nejadghaderi SA, Neupane Kandel S, Ngwa CH, Niazi RK, Nogueira de Sá AT, Noroozi N, Nouraei H, Nowroozi A, Nuñez-Samudio V, Nutor JJ, Nzoputam CI, Nzoputam OJ, Oancea B, Obaidur RM, Ojha VA, Okekunle AP, Okonji OC, Olagunju AT, Olusanya BO, Omar Bali A, Omer E, Otstavnov N, Oumer B, P a M, Padubidri JR, Pakshir K, Palicz T, Pana A, Pardhan S, Paredes JL, Parekh U, Park E-C, Park S, Pathak A, Paudel R, Paudel U, Pawar S, Pazoki Toroudi H, Peng M, Pensato U, Pepito VCF, Pereira M, Peres MFP, Perico N, Petcu I-R, Piracha ZZ, Podder I, Pokhrel N, Poluru R, Postma MJ, Pourtaheri N, Prashant A, Qattea I, Rabiee M, Rabiee N, Radfar A, Raeghi S, Rafiei S, Raghav PR, Rahbarnia L, Rahimi-Movaghar V, Rahman M, Rahman MA, Rahmani AM, Rahmanian V, Ram P, Ranjha MMAN, Rao SJ, Rashidi M-M, Rasul A, Ratan ZA, Rawaf S, Rawassizadeh R, Razeghinia MS, Redwan EMM, Regasa MT, Remuzzi G, Reta MA, Rezaei N, Rezapour A, Riad A, Ripon RK, Rudd KE, Saddik B, Sadeghian S, Saeed U, Safaei M, Safary A, Safi SZ, Sahebazzamani M, Sahebkar A, Sahoo H, Salahi S, Salahi S, Salari H, Salehi S, Samadi Kafil H, Samy AM, Sanadgol N, Sankararaman S, Sanmarchi F, Sathian B, Sawhney M, Saya GK, Senthilkumaran S, Seylani A, Shah PA, Shaikh MA, Shaker E, Shakhmardanov MZ, Sharew MM, Sharifi-Razavi A, Sharma P, Sheikhi RA, Sheikhy A, Shetty PH, Shigematsu M, Shin JI, Shirzad-Aski H, Shivakumar KM, Shobeiri P, Shorofi SA, Shrestha S, Sibhat MM, Sidemo NB, Sikder MK, Silva LMLR, Singh JA, Singh P, Singh S, Siraj MS, Siwal SS, Skryabin VY, Skryabina AA, Socea B, Solomon DD, Song Y, Sreeramareddy CT, Suleman M, Abdulkader R. Suliankatchi, Sultana S, Szócska M, Tabatabaeizadeh S-A, Tabish M, Taheri M, Taki E, Tan K-K, Tandukar S, Tat NY, Tat VY, Tefera BN, Tefera YM, Temesgen G, Temsah M-H, Tharwat S, Thiyagarajan A, Tleyjeh II, Troeger CE, Umapathi KK, Upadhyay E, Valadan Tahbaz S, Valdez PR, Van den Eynde J, van Doorn HR, Vaziri S, Verras G-I, Viswanathan H, Vo B, Waris A, Wassie GT, Wickramasinghe ND, Yaghoubi S, Yahya GATY, Yahyazadeh Jabbari SH, Yigit A, Yiğit V, Yon DK, Yonemoto N, Zahir M, Zaman BA, Zaman SB, Zangiabadian M, Zare I, Zastrozhin MS, Zhang Z-J, Zheng P, Zhong C, Zoladl M, Zumla A, Hay SI, Dolecek C, Sartorius B, Murray CJL, Naghavi M, Global mortality associated with 33 bacterial pathogens in 2019: a systematic analysis for the Global Burden of Disease Study 2019, The Lancet. 400 (2022) 2221–2248. 10.1016/S0140-6736(22)02185-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Pulingam T, Parumasivam T, Gazzali AM, Sulaiman AM, Chee JY, Lakshmanan M, Chin CF, Sudesh K, Antimicrobial resistance: Prevalence, economic burden, mechanisms of resistance and strategies to overcome, Eur. J. Pharm. Sci. 170 (2022) 106103. 10.1016/j.ejps.2021.106103. [DOI] [PubMed] [Google Scholar]
  • [3].Doualeh M, Payne M, Litton E, Raby E, Currie A, Molecular Methodologies for Improved Polymicrobial Sepsis Diagnosis, Int. J. Mol. Sci. 23 (2022) 4484. 10.3390/ijms23094484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Fukushima S, Hagiya H, Fujita K, Kamiyama S, Yamada H, Kishida M, Otsuka F, Clinical and microbiological characteristics of polymicrobial bacteremia: a retrospective, multicenter study, Infection. 50 (2022) 1233–1242. 10.1007/s15010-022-01799-7. [DOI] [PubMed] [Google Scholar]
  • [5].Armbruster CE, Smith SN, Johnson AO, DeOrnellas V, Eaton KA, Yep A, Mody L, Wu W, Mobley HLT, The Pathogenic Potential of Proteus mirabilis Is Enhanced by Other Uropathogens during Polymicrobial Urinary Tract Infection, Infect. Immun. 85 (2017) 10.1128/iai.00808-16. 10.1128/iai.00808-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Croxall G, Weston V, Joseph S, Manning G, Cheetham P, McNally A, Increased human pathogenic potential of Escherichia coli from polymicrobial urinary tract infections in comparison to isolates from monomicrobial culture samples, J. Med. Microbiol. 60 (2011) 102–109. 10.1099/jmm.0.020602-0. [DOI] [PubMed] [Google Scholar]
  • [7].Bertesteanu S, Triaridis S, Stankovic M, Lazar V, Chifiriuc MC, Vlad M, Grigore R, Polymicrobial wound infections: Pathophysiology and current therapeutic approaches, Int. J. Pharm. 463 (2014) 119–126. 10.1016/j.ijpharm.2013.12.012. [DOI] [PubMed] [Google Scholar]
  • [8].Bowler PG, Duerden BI, Armstrong DG, Wound Microbiology and Associated Approaches to Wound Management, Clin. Microbiol. Rev. 14 (2001) 244–269. 10.1128/cmr.14.2.244-269.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Stacy A, McNally L, Darch SE, Brown SP, Whiteley M, The biogeography of polymicrobial infection, Nat. Rev. Microbiol. 14 (2016) 93–105. 10.1038/nrmicro.2015.8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [10].Griffiths EC, Pedersen AB, Fenton A, Petchey OL, The nature and consequences of coinfection in humans, J. Infect. 63 (2011) 200–206. 10.1016/j.jinf.2011.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [11].Tay WH, Chong KKL, Kline KA, Polymicrobial–Host Interactions during Infection, J. Mol. Biol. 428 (2016) 3355–3371. 10.1016/j.jmb.2016.05.006. [DOI] [PubMed] [Google Scholar]
  • [12].Brogden KA, Guthmiller JM, Taylor CE, Human polymicrobial infections, The Lancet. 365 (2005) 253–255. 10.1016/S0140-6736(05)17745-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Short FL, Murdoch SL, Ryan RP, Polybacterial human disease: the ills of social networking, Trends Microbiol. 22 (2014) 508–516. 10.1016/j.tim.2014.05.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [14].Seymour CW, Gesten F, Prescott HC, Friedrich ME, Iwashyna TJ, Phillips GS, Lemeshow S, Osborn T, Terry KM, Levy MM, Time to Treatment and Mortality during Mandated Emergency Care for Sepsis, N. Engl. J. Med. 376 (2017) 2235–2244. 10.1056/NEJMoa1703058. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Bassetti M, Rello J, Blasi F, Goossens H, Sotgiu G, Tavoschi L, Zasowski EJ, Arber MR, McCool R, Patterson JV, Longshaw CM, Lopes S, Manissero D, Nguyen ST, Tone K, Aliberti S, Systematic review of the impact of appropriate versus inappropriate initial antibiotic therapy on outcomes of patients with severe bacterial infections, Int. J. Antimicrob. Agents. 56 (2020) 106184. 10.1016/j.ijantimicag.2020.106184. [DOI] [PubMed] [Google Scholar]
  • [16].Zasowski EJ, Bassetti M, Blasi F, Goossens H, Rello J, Sotgiu G, Tavoschi L, Arber MR, McCool R, Patterson JV, Longshaw CM, Lopes S, Manissero D, Nguyen ST, Tone K, Aliberti S, A Systematic Review of the Effect of Delayed Appropriate Antibiotic Treatment on the Outcomes of Patients With Severe Bacterial Infections, Chest. 158 (2020) 929–938. 10.1016/j.chest.2020.03.087. [DOI] [PubMed] [Google Scholar]
  • [17].Scerbo MH, Kaplan HB, Dua A, Litwin DB, Ambrose CG, Moore LJ, Murray CCK, Wade CE, Holcomb JB, Beyond Blood Culture and Gram Stain Analysis: A Review of Molecular Techniques for the Early Detection of Bacteremia in Surgical Patients, Surg. Infect. 17 (2016) 294–302. 10.1089/sur.2015.099. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].Lagier J-C, Edouard S, Pagnier I, Mediannikov O, Drancourt M, Raoult D, Current and Past Strategies for Bacterial Culture in Clinical Microbiology, Clin. Microbiol. Rev. 28 (2015) 208–236. 10.1128/CMR.00110-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Maugeri G, Lychko I, Sobral R, Roque ACA, Identification and Antibiotic-Susceptibility Profiling of Infectious Bacterial Agents: A Review of Current and Future Trends, Biotechnol. J. 14 (2019) 1700750. 10.1002/biot.201700750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Hu L, Han B, Tong Q, Xiao H, Cao D, Detection of Eight Respiratory Bacterial Pathogens Based on Multiplex Real-Time PCR with Fluorescence Melting Curve Analysis, Can. J. Infect. Dis. Med. Microbiol. 2020 (2020) e2697230. 10.1155/2020/2697230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [21].Bender AC, Faulkner JA, Tulimieri K, Boise TH, Elkins KM, High Resolution Melt Assays to Detect and Identify Vibrio parahaemolyticus, Bacillus cereus, Escherichia coli, and Clostridioides difficile Bacteria, Microorganisms. 8 (2020) 561. 10.3390/microorganisms8040561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [22].Athamanolap P, Hsieh K, Chen L, Yang S, Wang T-H, Integrated Bacterial Identification and Antimicrobial Susceptibility Testing Using PCR and High-Resolution Melt, Anal. Chem. 89 (2017) 11529–11536. 10.1021/acs.analchem.7b02809. [DOI] [PubMed] [Google Scholar]
  • [23].Velez DO, Mack H, Jupe J, Hawker S, Kulkarni N, Hedayatnia B, Zhang Y, Lawrence S, Fraley SI, Massively parallel digital high resolution melt for rapid and absolutely quantitative sequence profiling, Sci. Rep. 7 (2017) 42326. 10.1038/srep42326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [24].Athamanolap P, Hsieh K, O’Keefe CM, Zhang Y, Yang S, Wang T-H, Nanoarray Digital Polymerase Chain Reaction with High-Resolution Melt for Enabling Broad Bacteria Identification and Pheno–Molecular Antimicrobial Susceptibility Test, Anal. Chem. 91 (2019) 12784–12792. 10.1021/acs.analchem.9b02344. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Gaddes DE, Lee P, Trick AY, Athamanolap P, O’Keefe CM, Puleo C, Hsieh K, Wang T-H, Facile Coupling of Droplet Magnetofluidic-Enabled Automated Sample Preparation for Digital Nucleic Acid Amplification Testing and Analysis, Anal. Chem. 92 (2020) 13254–13261. 10.1021/acs.analchem.0c02454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [26].Sinha M, Mack H, Coleman TP, Fraley SI, A High-Resolution Digital DNA Melting Platform for Robust Sequence Profiling and Enhanced Genotype Discrimination, SLAS Technol. Transl. Life Sci. Innov. 23 (2018) 580–591. 10.1177/2472630318769846. [DOI] [PubMed] [Google Scholar]
  • [27].Langouche L, Aralar A, Sinha M, Lawrence SM, Fraley SI, Coleman TP, Data-driven noise modeling of digital DNA melting analysis enables prediction of sequence discriminating power, Bioinformatics. 36 (2020) 5337–5343. 10.1093/bioinformatics/btaa1053. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [28].Gundry CN, Dobrowolski SF, Martin YR, Robbins TC, Nay LM, Boyd N, Coyne T, Wall MD, Wittwer CT, Teng DH-F, Base-pair neutral homozygotes can be discriminated by calibrated high-resolution melting of small amplicons, Nucleic Acids Res. 36 (2008) 3401–3408. 10.1093/nar/gkn204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [29].Athamanolap P, Shin DJ, Wang T-H, Droplet Array Platform for High-Resolution Melt Analysis of DNA Methylation Density, J. Lab. Autom. 19 (2014) 304–312. 10.1177/2211068213507923. [DOI] [PubMed] [Google Scholar]
  • [30].Fraley SI, Athamanolap P, Masek BJ, Hardick J, Carroll KC, Hsieh Y-H, Rothman RE, Gaydos CA, Wang T-H, Yang S, Nested Machine Learning Facilitates Increased Sequence Content for Large-Scale Automated High Resolution Melt Genotyping, Sci. Rep. 6 (2016) 19218. 10.1038/srep19218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Andini N, Wang B, Athamanolap P, Hardick J, Masek BJ, Thair S, Hu A, Avornu G, Peterson S, Cogill S, Rothman RE, Carroll KC, Gaydos CA, Wang JT-H, Batzoglou S, Yang S, Microbial Typing by Machine Learned DNA Melt Signatures, Sci. Rep. 7 (2017) 42097. 10.1038/srep42097. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Fraley SI, Hardick J, Jo Masek B, Athamanolap P, Rothman RE, Gaydos CA, Carroll KC, Wakefield T, Wang T-H, Yang S, Universal digital high-resolution melt: a novel approach to broad-based profiling of heterogeneous biological samples, Nucleic Acids Res. 41 (2013) e175. 10.1093/nar/gkt684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Yu J, Zhou T, Zhu B, Wei Y, Li X, Liu Y, Species-Specific Identification of Streptococcus based on DNA Marker in 16S–23S rDNA Internal Transcribed Spacer, Curr. Microbiol. 77 (2020) 1569–1579. 10.1007/s00284-020-01975-8. [DOI] [PubMed] [Google Scholar]
  • [34].Milani C, Alessandri G, Mangifesta M, Mancabelli L, Lugli GA, Fontana F, Longhi G, Anzalone R, Viappiani A, Duranti S, Turroni F, Costi R, Annicchiarico A, Morini A, Sarli L, Ossiprandi MC, van Sinderen D, Ventura M, Untangling Species-Level Composition of Complex Bacterial Communities through a Novel Metagenomic Approach, mSystems. 5 (2020) e00404–20. 10.1128/mSystems.00404-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Fida M, Khalil S, Abu Saleh O, Challener DW, Sohail MR, Yang JN, Pritt BS, Schuetz AN, Patel R, Diagnostic Value of 16S Ribosomal RNA Gene Polymerase Chain Reaction/Sanger Sequencing in Clinical Practice, Clin. Infect. Dis. 73 (2021) 961–968. 10.1093/cid/ciab167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Church DL, Cerutti L, Gürtler A, Griener T, Zelazny A, Emler S, Performance and Application of 16S rRNA Gene Cycle Sequencing for Routine Identification of Bacteria in the Clinical Microbiology Laboratory, Clin. Microbiol. Rev. 33 (2020) 10.1128/cmr.00053-19. 10.1128/cmr.00053-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [37].Winand R, Bogaerts B, Hoffman S, Lefevre L, Delvoye M, Van Braekel J, Fu Q, Roosens NH, De Keersmaecker SC, Vanneste K, Targeting the 16S rRNA Gene for Bacterial Identification in Complex Mixed Samples: Comparative Evaluation of Second (Illumina) and Third (Oxford Nanopore Technologies) Generation Sequencing Technologies, Int. J. Mol. Sci. 21 (2020) 298. 10.3390/ijms21010298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Lee P-W, Chen L, Hsieh K, Traylor A, Wang T-H, Harnessing Diversities in Digital Melt Curves for Accurate Identification of Bacteria, Anal. Chem. 95 (2023) 15522–15530. 10.1021/acs.analchem.3c01654. [DOI] [PubMed] [Google Scholar]
  • [39].Andini N, Hu A, Zhou L, Cogill S, Wang T-H, Wittwer CT, Yang S, A “Culture” Shift: Broad Bacterial Detection, Identification, and Antimicrobial Susceptibility Testing Directly from Whole Blood, Clin. Chem. 64 (2018) 1453. 10.1373/clinchem.2018.290189. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Reed GH, Kent JO, Wittwer CT, High-resolution DNA melting analysis for simple and efficient molecular diagnostics, Pharmacogenomics. 8 (2007) 597–608. 10.2217/14622416.8.6.597. [DOI] [PubMed] [Google Scholar]
  • [41].Athamanolap P, Parekh V, Fraley SI, Agarwal V, Shin DJ, Jacobs MA, Wang T-H, Yang S, Trainable High Resolution Melt Curve Machine Learning Classifier for Large-Scale Reliable Genotyping of Sequence Variants, PLOS ONE. 9 (2014) e109094. 10.1371/journal.pone.0109094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Founou RC, Founou LL, Essack SY, Clinical and economic impact of antibiotic resistance in developing countries: A systematic review and meta-analysis, PLOS ONE. 12 (2017) e0189621. 10.1371/journal.pone.0189621. [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

1
NIHMS1970192-supplement-1.docx (1.1MB, docx)

RESOURCES