Two-Step Reverse Transcription Droplet Digital PCR Protocols for SARS-CoV-2 Detection and Quantification

Polymerase chain reaction (PCR), a well-recognized technique, has undergone several transformations since its advent to become a powerful technique capable of providing answers to nucleic acid research. These transformations have been a constant improvement of the old technique. These transformations can be summarized into three generations¹. The first generation is conventional PCR that relies on gel electrophoresis to quantify and detect amplified targets. The second generation is quantitative real time PCR (qPCR) that can detect samples in real time and rely on a standard curve to directly quantify targets in a sample. The third generation, digital PCR (dPCR), can perform both detection, and absolute quantification of nucleic acid targets without the need of a standard curve. dPCR has also been improved further from reaction chambers being separated by the wells of a wall into emulsions of oil, water, and stabilizing chemicals within the same well as seen in droplet-based digital PCR². In droplet digital PCR (ddPCR), a sample is partitioned into thousands of nanoliter-sized droplets containing individual targets that will later be quantified using Poisson statistics^2,3,4. This technique gives ddPCR an edge in quantifying low abundant targets when compared to the other generations of PCR.

Recently, multiple applications have highlighted the superiority of ddPCR over the commonly used qPCR when detecting and quantifying low abundant targets^1,5,6. SARS-CoV-2 is no exception to these applications^{7,8,9,10,11,12}. Since the outbreak of SARS-CoV-2, scientists have been working on all fronts to come up with solutions on how to diagnose the virus and detect it efficiently. The current gold standard still remains to be qPCR¹³. However, RT-ddPCR has been shown to be more accurate in detecting low abundant SARS-CoV-2 targets from both environmental and clinical samples when compared to RT-qPCR^{7,8,9,10,11,12}. Most of the SARS-CoV-2 ddPCR published works depend on simplex assays with the multiplex ones depending on commercial assays. Hence, more should be done to explain how to develop multiplex RT-dPCR assays for SARS-CoV-2 detection.

In a proper assay design, multiplexing can be used to save on cost, increase sample throughput, and maximize on the number of targets that can be sensitively detected within a small sample. When multiplexing with ddPCR, one must take account of how many fluorophores can be detected in a particular system. Some ddPCR platforms can support up to three channels while others support only two channels. Hence, when multiplexing with two channels, one has to use different approaches, including higher order multiplexing to detect more than two targets^14,15,16. In this work, a two color ddPCR detection system is used to show steps on how to develop different SARS-CoV-2 RT-ddPCR assays that can be adapted for different research applications.

Ethical statement
Wuhan Institute of Virology (WHIOV) is among the labs and institutes approved by China CDC of Wuhan city to conduct research on SARS-CoV-2 and detect COVID-19 from clinical samples. Research on developing new diagnostic techniques for COVID-19 using clinical samples has also been approved by the ethical committee of Wuhan Institute of Virology (2020FCA001).

1. Sample processing workflow (Figure 1A)

NOTE: Throughout the protocol, it is important to use separate rooms with dedicated pipettes for sample handling (extraction and storage), reagent/mastermix preparation and storage, reaction mix preparation (sample plus mastermix), and detection, to avoid cross contamination. The assays to be developed can be used in the detection of clinical samples or research samples. All samples should be treated as if they can transmit infectious agents even when using safe laboratory procedures. Sample processing steps should be done in a biosafety level 2 (BSL-2) laboratory following strict BSL-2 rules, including wearing of appropriate personal protective equipment (PPE).

1. Sample inactivation
   1. Take 1 mL of SARS-CoV-2 sample to a BSL-2 laboratory. Heat-inactivate samples at 65 °C for 30 min.
      NOTE: Samples may come from various sources, hence, one should make sure the volume to be inactivated is sufficient to ensure subsequent RNA extraction. Most extraction kits require a small sample volume of up to 200 µL, hence, a sample volume of about 1 mL would be sufficient for inactivation. Surfactants, kits, and lysis buffers may also be used for direct inactivation and RNA extraction according to the labs own SOP.
   2. Take samples to a biosafety cabinet (BSC) and let them stand at room temperature for 10 min to allow potential aerosols to settle. Store samples in 4 °C for up to 24 h if not processed immediately.
      NOTE: No specific container is required for sample storage. Samples may be stored in the tubes or containers used during inactivation. For this work, most samples were stored in viral transport media (VTM) tubes or 1.5 mL tubes.
2. RNA extraction
   NOTE: Many RNA extraction instruments and kits with specific protocols are available for ready use. Here, an automated procedure following the manufacturer's instructions is presented.
   1. Take out the pre-filled 96-deep well orifice plate; invert it upside-down gently to mix the magnetic beads. After mixing, gently shake the plate on a flat surface to concentrate reagent and magnetic beads that may be on the wall to the plate's bottom.
   2. Carefully tear off the aluminum foil sealing film to avoid vibration of the plate and liquid spillage. Add 20 µL of proteinase K and 200 µL of the sample per well in rows 2 and 8 of the 96-well plate. Then, place the 96-well plate on the base of the 32-channel automatic nucleic acid extraction instrument.
      NOTE: Run the program on the computer within 1 h after adding the sample.
   3. Insert the magnetic bar sleeve into the card slot of the 32-channel automatic nucleic acid extraction instrument. Select the program GF-FM502T5-TR"1"GF-FM502T5YH (quick version) and run it.
   4. After completion of the extraction, take out the nucleic acid samples in rows 6 and 12 (about 100 µL/sample) and distribute in a clean nuclease-free 96-well plate. Store the samples at 4 °C for up to 24 h until use or at -20 °C for a longer time.
      NOTE: It is recommended to use a 100 µL multichannel pipette to transfer samples into a new 100-200 µL plate as it can directly match the columns of the sample plate.
3. cDNA generation
   NOTE: For samples stored for a long time, avoid multiple freeze/thaw cycles. Perform cDNA generation using a cDNA kit following the manufacturer's instructions.
   1. In a 100 or 200 µL PCR tube placed on ice/cooling block, add 2 µL of 5x RT master mix (e.g., Perfect Real Time), 5 µL of sample RNA, and 3 µL of RNase free ddH2O per reaction (total volume 10 µL).
   2. Place the PCR tube in a thermal cycler set to run at 37 °C for 15 min (reverse-transcription), 85 °C for 5 s (heat inactivation of reverse transcriptase), and 4 °C for an infinite time.
      ​NOTE: Process the samples immediately, or store at 4 °C for up to 24 h until use or at -20 °C for up to 6 months.

2. Optimization of ddPCR assay and workflow

NOTE: Optimize the assays before/after reading the droplets. Dependent on the results, they can be optimized at any point of the work to achieve better results. Below are some common factors to be considered when optimizing ddPCR experiments.

1. Validate the ddPCR primers and probes (Table 1).
   NOTE: The primers ORF1ab and N were adapted from China CDC¹⁷, the human endogenous control gene (RPP30) from Lu et al.¹⁸, and RBD2 from Nyaruaba et al.¹³. All primers and probes except RBD2 had already been ddPCR tested and optimized^7,8,18.
2. Run the ddPCR test sample controls comprising a positive control (a sample with all four targets), No Template Control (NTC; Nuclease free water or sample with no targets), and extraction/negative control (total nucleic acid extracted from a non-infectious cultured human cell or pooled samples from health volunteers).
   NOTE: Standard control samples with known copies of target genes are preferred. However, in the absence of standards, use the available samples to define their controls or sample matrix. Run the controls together with test samples.
3. Optimize the annealing temperature (Figure 2D) by inserting a temperature gradient of 55 °C to 65 °C in the annealing step of PCR (step 3.2).
   NOTE: Annealing temperature optimization helps to find the best temperature where droplet separation is maximum (easy target identification) with minimal rain. After obtaining the results, narrow down the temperature range for better optimization, e.g., 55 °C to 60 °C.
4. Optimize the primer and probe concentrations. If optimum separation between positive and negative droplets is not achieved in the assays, vary the primer (300-900 nM) and/or probe (100-400 nM) concentrations.
   NOTE: Lowering primer/probe concentrations results in a lower target droplet amplitude and increasing it also increases the target's amplitude. This can be helpful in amplitude-based multiplexing.
5. Test the analytical sensitivity considering different parameters including, limit of blank (LoB), limit of detection (LoD), specificity, and sensitivity using both standard and test samples, as preferred.

3. ddPCR workflow (Figure 1B) and assay development (Table 2)

NOTE: Like other ddPCR detection systems, this workflow also consists of four steps (Figure 1B), including reaction mix preparation, droplet generation, PCR amplification, and droplet reading.

1. Reaction mix preparation
   NOTE: Prepare all assays in a clean separate room and inside a BSC. Observe standard precautions, including use of clean gloves, clean pipettes, nuclease free water, and disinfectants. Aliquot reagents in separate tubes and store at -20 ^oC to avoid repeated freeze thawing cycles. Prepare assays as shown in Table 2. Primer and probe stock solutions should be stored in high concentrations of 100 µM. The working solutions can be stored in aliquots of 20-40 µM (diluted in nuclease free water).
   1. Common steps in all assays
      NOTE: Before preparing the assays, calculate the total volume of mastermix needed based on the number of samples. Here, each mastermix component volume was multiplied by 1.05 to account for any pipetting errors.
      1. Thaw and equilibrate reaction materials to room temperature. Vortex the reactions components briefly (30 s) to ensure homogeneity, and briefly centrifuge to collect the contents at the bottom of the tube.
      2. Prepare a primer-probe (PP) mix per assay by mixing specific reaction volumes as detailed in Table 2 from the working solutions.
      3. Prepare the mastermix by mixing 11 µL (1x) of 2x ddPCR supermix for probes (No dUTP), PP mix (dependent on assay as shown in Table 2 from the working solution) and nuclease free water to a final volume of 19.8 µL per well. Distribute the mastermix into the wells of a nuclease-free 96-well ddPCR plate based on the number of samples.
         ​NOTE: For the sample plate, make sure all 8 wells in one column have the solution. If only a few wells are used, e.g., 5 samples, fill the other 3 wells with 22 µL of nuclease free water or control buffer each.
      4. To get the final reaction mix volume (22 µL), add 2.2 µL of cDNA sample per well containing mastermix. Seal the plate with a disposable PCR plate sealer.
         NOTE: Perform the sample additions in the designated room. Include the controls i.e., positive, NTC, and extraction. Additionally, if the RNA samples are of low concentration, reduce the volume of water in the mastermix and increase the sample volume accordingly up to 5.5 uL.
      5. Once the reaction mix is distributed per well, vortex briefly (15-30 s), and centrifuge (10-15 s) to collect the contents at the bottom of the plate. Proceed for droplet generation.
2. Automated droplet generation (Supplementary Figure 1)
   NOTE: Different droplet generators can be used; however, this study was performed with an automated droplet generator (AutoDG). To avoid amplicon contamination, ensure the droplet generator and readers have dedicated space in separate areas. When loading consumables, it is recommended to start loading from the back of the instrument working toward the front to avoid moving hands over the consumables to avoid cross contamination.
   1. Obtain all consumables needed to set up the AutoDG, including AutoDG oil for probes, DG32 Cartridges, pipette tips, cooling block, sample plate, droplet plate (new clean ddPCR plate), and a waste bin.
   2. On the AutoDG touch screen, touch Configure Sample Plate and select columns where samples are located on sample plate and press OK.
      NOTE: The screen will turn yellow indicating LOAD where consumables should be loaded.
   3. Open the AutoDG door and load consumables in their respective places.
      NOTE: The respective places where consumables should be loaded will have a yellow indicator that will later turn green if the consumables are loaded. This is similar to the touchscreen.
   4. Load the sample plate and droplet generation plate.
      NOTE: The droplet generation plate should be placed on the cooling block. Ensure the cooling block is purple (ready to use) and not pink (should be kept in a freezer until the color turns back to purple).
   5. Close the AutoDG door, ensure the screen is green (consumables loaded), and select/ensure the oil type is Oil for Probes before pressing Start Droplet Generation.
   6. Wait for the droplets to be generated. Open the door once the screen shows Droplets Ready and remove the plate containing the droplets.
   7. Seal the plate with a pierceable foil seal using a plate sealer set to run at 185 ^oC for 5 s. Proceed to PCR amplification.
      NOTE: PCR amplification should be started within 30 min after droplet generation.
3. PCR amplification
   1. Insert the sealed droplet plate into a 96-deep well thermal cycler, set the sample volume to 40 µL and lid temperature to 105 °C.
   2. PCR amplify the droplets using the program: 95 °C for 10 min (enzyme activation), 40 cycles of denaturation at 94 °C for 30 s and 1 min annealing/extension at 57 °C, enzyme deactivation at 98 °C for 10 min, and holding step at 4 °C indefinitely. After the PCR, read the droplets or store at 4 °C for up to 24 h.
      NOTE: Use a ramp rate of 2 °C /s at all steps, as ramp rates may differ for different thermal cyclers. Hold the plate for at least 30 min at 4 °C to allow droplet stabilization before reading the droplets.
4. Droplet reading
   1. Transfer the thermal cycled 96-well plate into a droplet reader, and on a computer connected to the droplet reader, open/start the accompanying software.
   2. In the setup mode, select New and then double-click on any well to open the Well Editor dialog box. Select the wells to be read and choose Experiment: ABS, Supermix: ddPCR Supermix for Probes (no dUTP), Target 1 Type: Ch1 Unknown, Target 2 Type: Ch2 Unknown.
      NOTE: Negative, blank, NTC, or positive can be selected from the target 1/2 type dropdown menu for the control samples.
   3. Assign the target or sample names based on plate layout and select Apply. When done, select OK, and save the created template. Click on Run and when prompted choose the right color (FAM/HEX).
      NOTE: Priming the system before running is recommended before the first run of the day. Also, check whether all lights in the reader are green and if not, follow the recommended application on the tool tip status.
   4. After data acquisition is finished, remove the plate from the reader. Analyze data as required.
      NOTE: Since preliminary results can be seen on the screen, running the positive samples in the first wells can be used for real-time quality check.

4. Data analysis (Supplementary Figures 2 and 3)

1. Check the data from all the wells for total number of droplets. If the droplet count is <10,000, then discard the results and repeat the assay. Set a cut off to accept positive and negative results. E.g., A droplet count of up to 3 or more positive droplets can be considered as the cut off for positive results.
   NOTE: Data for all assays, including simplex, duplex, triplex probe mix, and quadruplex assays are generated using the accompanying software. However, this software is only suitable for simplex and duplex assays. For high order multiplex assays (>3 targets), an external software is needed.
2. Simplex and duplex assays
   NOTE: The external software can also analyze simplex and duplex data. However, using the accompanying software is easier for these assays, hence used in this article.
   1. Double click on the .qlp file to be analyzed to open it and select Analyze.
   2. Use the threshold tools in the 1D Amplitude and 2D Amplitude to distinguish between positive and negative droplets for each well in the correct channel. The NTC and positive control samples can be used as a guidance for threshold setting.
      ​NOTE: FAM results are seen in Channel 1 while HEX/VIC in channel 2.
   3. After threshold setting, the results can be exported as a .csv file and further analyzed in Excel or recorded by reading directly on the results window.
3. Triplex probe mix (Supplementary Figure 2) and quadruplex assays (Supplementary Figure 2)
   NOTE: Before analyzing the triplex probe mix and the quadruplex assay, download the external software. Refer to the minimal system requirements before installing.
   1. Open the .qlp file by right clicking on it and choosing Open with the external software, or by opening the external software and clicking on the Browse option to locate the .qlp file in your folder, or by simply dragging the .qlp file and dropping it on the already open external software.
   2. In the plate editor tab, select the wells to be analyzed on the right side of the tab.
      1. For triplex probe mix, select Direct Quantification (DQ) as experiment type, select Probe Mix Triplex as assay information and enter target names accordingly, and then click on Apply.
      2. For quadruplex, select Direct Quantification (DQ) as experiment type, select Amplitude Multiplex as assay information and enter target names accordingly, and then click on Apply.
   3. On the left side of the 2D Amplitude tab, use the Graph Tools to assign specific colors to different target clusters following the Select to Assign Cluster window pop up suggestions.
      NOTE: Preferred cluster mode can be applied based on the users' preference.
   4. Once the target colors are assigned, quantification data can be read in the Well Data Window on the lower right. Export data to Excel/csv for further analysis using the triple-bar icon on the upper right hand of the well data table.

In a proof-of-concept study, the multiplex assays analytical performance was tested on clinical and research samples¹⁹. The performance of the multiplex assays was superior to that of an RT-PCR¹⁹. Since low numbers of droplets may indicate a problem during droplet generation, in this article a cutoff of 10,000 droplets per well was set based on empirical data.

A good separation between positive and negative droplets with minimal rain interference can help in data analysis. Hence, in a good experiment, assay optimization is key¹⁹. As seen in the temperature gradient analysis results in Figure 2D, high annealing temperatures (e.g., 65 °C) could not clearly distinguish positive droplets from negative droplets when a duplex assay (N (FAM) and RPP30 (HEX)) was run. However, with a decrease in annealing temperature, optimal separation between positive and negative droplets was achieved. A temperature of 57 °C was found optimal. This can also be observed in other assays¹⁹.

Using a two color (FAM/HEX) RT-ddPCR detection system, it is possible to detect one (Figure 2A, B), two (Figure 2C), three (Figure 3A), and four (Figure 3B) SARS-CoV-2 targets within a single sample. During data analysis, the accompanying software used to read droplets can only analyze simplex and duplex assays as shown in Figure 2. This means that for higher order multiplex assays (>3 targets), an external software should be used for data analysis as shown in Figures 3, S2, and S3. It is important to also note that the external software can also be used to analyze simplex and duplex data. For simplex and duplex data, analysis is quite simple as targets are separated as either positive or negative droplets in their respective channels as shown in Figure 2. A NTC sample can help in the location of negative droplets that can in turn help one set thresholds for data analysis as shown in Figure 2A. For the duplex assay, analysis can be done in individual channels (Figure 2B i,ii) or in the 2D Amplitude (Figure 2B iii).

For higher order multiplex assays, data analysis is not straightforward, and attention should be focused on droplet target assignment. After installing the external software, select wells to be analyzed, select the appropriate experiment type, and assign target clusters based on experiment type as shown in Figure S2 and S3. The Select to Assign Cluster window pop will guide one on how to assign clusters in the higher multiplex assays. Use the graph tools to assign up to 8 droplet clusters for the triplex probe mix assay (Figure 3A), and up to 16 droplet clusters for the quadruplex amplitude-based assay (Figure 3B).

After assigning clusters and thresholds, quantification data for each target in the form of copies/µL can be read in the well data window on the lower right of the external software. This data can be used to estimate the number of copies of targets in the starting sample. E.g., if 2.2 µL of the sample was used in a final volume of 22 µL, and the software recorded 30.5 copies/µL for ORF1ab, there were 30.5 x 22 = 671 copies of ORF1ab in the PCR mix. The mix contained 2.2 µL of original sample, hence, there were 671 copies of ORF1ab in the starting sample, and 671/2.2 = 305 copies/µL of ORF1ab in the original sample. This method can be used to estimate the concentration of each target in all assays.

Table 1: Primer and probe sequences used to develop the different SARS-CoV-2 assays.

Table 2: Final concentration of different primer and probe pairs per assay.

Figure 1: Sample processing and droplet digital PCR workflow. (A) The sample processing workflow includes sample collection and transport to a BSL-2 facility, inactivation, extraction, and cDNA generation. (B) The droplet digital PCR workflow begins with preparation of the mastermix, loading the mastermix into a ddPCR plate and adding sample(s), generating droplets, amplifying targets inside droplets by PCR, and finally reading the amplified droplets using a droplet reader. Please click here to view a larger version of this figure.

Figure 2: Simplex and duplex assay results, including annealing temperature optimization. (A) Simplex assay results when a single target (RBD2) was FAM labeled. (B) Simplex assay results when a single target (RPP30) was HEX labeled. (C) Duplex assay result of two targets in 1D and 2D (iii) Channels after N (i) was FAM labeled and RPP30 (ii) was HEX labeled. (D) Annealing temperature gradient (65 °C to 55 °C) results of a duplex assay labeled with ORF1ab (FAM) and RPP30 (HEX). Please click here to view a larger version of this figure.

Figure 3: Triplex probe mix and quadruplex amplitude-based multiplex assay results. (A) Triplex probe mix assay results when three targets were labeled with the following ratios of FAM:HEX; RBD2 (1:0), N (0.5:0.5), and RPP30 (0:1). (B) Quadruplex amplitude-based assay results after four targets were labeled with the following ratios of FAM:HEX; RBD2(0.5:0), N (1:0), RPP30 (0:0.5), and ORF1ab (0:1). 1 and 0.5 are probe concentrations 250 nM and 125 nM, respectively. Please click here to view a larger version of this figure.

Supplementary Figure 1: Droplet generation using an automated droplet generator. (A) Consumables needed to set up the AutoDG. (B) On the AutoDG touch screen, touch Configure Sample Plate and select columns where samples are located on the sample plate and press OK. (C) Once selected, the screen turns yellow indicating where consumables should be loaded. (D) Open the AutoDG door and load consumables in the respective places. Load consumables from the back working toward the front. Make sure each time a consumable is added the light turns from yellow to green at that location. (E) Ensure the oil type used is oil for probes. (F) Close the AutoDG door and ensure all reagents are set in place by checking whether the AutoDG touch screen is green. (G) Press START Droplet Generation to generate droplets. (H) After droplet generation is complete, open the AutoDG and remove the droplet plate. (I) Seal the droplet plate using a pierceable foil heat seal. Please click here to download this File.

Supplementary Figure 2: Steps on analysis of amplitude-based multiplex ddPCR assay results using an external software. (A) Install the external software on a computer. (B) Open the .qlp file by either right clicking on it and choosing Open with the installed external software or opening the external software and clicking on the Browse option to locate the file in your folder. Alternatively, one can drag the .qlp file and drop it the open external software to open it. (C) Once open, on the right side of the Plate Editor tab, choose probe mix triplex from the drop-down menu and assign target information accordingly, and click on Apply. (D) On the left side of the 2D amplitude tab, use the Graph Tools to assign specific colors to different targets for detection and quantification. Once droplet cluster targets are identified, the quantification results can be seen on the Well Data window in the same 2D amplitude tab. Please click here to download this File.

Supplementary Figure 3: Steps on analysis of amplitude based multiplex ddPCR assay results using an external software. (A) Install the external software on a computer. (B) Open the .qlp file by either right clicking on it and choosing Open with the installed external software, or opening the external software and clicking on the Browse option to locate the file in your folder. Alternatively, one can drag the .qlp file and drop it in the already open external software to open it. (C) Once open, on the right side of the Plate Editor tab, choose amplitude multiplex from the drop-down menu and assign target information accordingly, and click on Apply. (D) On the left side of the 2D amplitude tab, use the Graph Tools to assign specific colors to different targets for detection and quantification. Once droplet cluster targets are identified, the quantification results can be seen on the Well Data window in the same 2D amplitude tab. Please click here to download this File.

Few resources are available on how to develop RT-ddPCR assays for SARS-CoV-2 detection. Though not used in this article, standard samples with known copies may be used to develop and optimize assays. In this work however, SARS-CoV-2 samples grown in Vero-E6 cells were spiked in a background of human genomic RNA and used as standard samples to develop the assays. Proper primer and probe sequences are essential when developing assays. Since most preliminary work on SARS-CoV-2 RT-ddPCR used the China CDC primer and probes targeting the ORF1ab and N gene, they had found them fit to be included in this work^7,8,9,10,11. The analytical specificity and sensitivity of these primers have also been compared using ddPCR in the previous work⁷. The RBD2 primers and probe¹³ were developed in-house and found fit for RT-ddPCR. Since the assays may be used for different applications, including diagnosis, the Ribonuclease P protein subunit p30 (RPP30) specific human gene was included in all multiplex assays. These gene can be used as a human endogenous control for diagnostic experiments. However, in the case of environmental sampling or other research that do not need the human reference gene, one may alternate this target with another SARS-CoV-2 target.

In all assays, it is important to include controls to validate assays and experimental data. These controls may include: NTC (nuclease free water as sample) to help in setting thresholds and locating negative droplet cluster; positive control (sample with all SARS-CoV-2 targets, including human reference gene) to assess reagent failure, primer and probe integrity, and substantial reverse transcription detection/location of positive droplet targets; and extraction controls (pooled human samples from healthy volunteers or total nucleic acid extracted from a noninfectious cultured human cell) to detect extraction step failure or success.

Most ddPCR systems, including the QX200 system have a narrow dynamic range from 1 to 120,000 copies/20 µL reaction. When detecting unknow samples, the target concentration in the starting sample is often unknown and this may pose challenges when quantifying highly concentrated samples. To overcome this, it is recommended that when quantifying samples suspected to contain high amounts of target molecules (such as cell culture), one should plan to reduce the starting sample accordingly. In the case where the target copy number/genome is unknown, one should determine the optimal starting amount through a series of four tenfold serial dilution of each sample at the expected digital range. By assaying these four points, it is ensured that one of the data points is within the optimal digital range.

Using high primer and low probe concentration is a perquisite of most simplex and duplex ddPCR experiments. This difference in concentration increases the amplitude separation distance between the positive and negative droplet clusters, hence making it easy to analyze data. However, when developing higher order multiplex assays, changes in these concentrations may lead to changes in the position of positive droplet targets as shown in Figure 3 and 4. As a result, one assay optimization option apart from annealing temperature would be to change target primer or probe concentration to distinguish droplets. This phenomenon has been used and explained before^14,15,16.

Despite the assay performance, this work is a two-step RT-ddPCR workflow. The extra reverse transcription step before ddPCR, gives room for contamination of samples. However, if careful and proper sample handling techniques are used, this will be a non-issue. Positively, DNA is known to be more stable than RNA. The conversion of RNA to cDNA may extend the sample's shelf life during storage as compared to RNA. A two-step RT-ddPCR experiment is also cheaper than a one-step RT-ddPCR experiment.

Compared to RT-qPCR, RT-ddPCR is expensive. Hence, considerations should be taken when performing RT-ddPCR. For example, during diagnosis, one may use RT-ddPCR in the case where their samples have low abundant targets. However, there is a possibility that the costs of dPCR instruments and reagents will drop soon, and the technique will be adapted in many laboratories, as it happened in the past with regular PCR and qPCR. Hence, it is important to set protocols such as this for current and future users of dPCR. In conclusion, the developed assays give room for prospective users to vary targets based on their applications. Multiplexing will ensure that one can efficiently detect many targets within a single sample in a single reaction. So far, this may be the first protocol that gives a full detail on how to use the AutoDG system, including the external software in SARS-CoV-2 detection. More work on assay optimization still needs to be done to achieve better separation in the quadruplex assay. The use of standards will also help improve the developed assays.