key: cord-0732267-7a5xegus authors: Azmi, I.; Faizan, M. I.; Kumar, R.; Yadav, S. R.; Chaudhary, N.; Singh, D. K.; Butola, R.; Ganotra, A.; Joshi, G. D.; Iqbal, J.; Joshi, M. C.; Ahmad, T. title: A saliva-based RNA extraction-free workflow integrated with Cas13a for SARS-CoV-2 detection date: 2020-11-10 journal: nan DOI: 10.1101/2020.11.07.20227082 sha: ca1d463a5da15f21c06d7ba8e3adacdc0074d148 doc_id: 732267 cord_uid: 7a5xegus A major bottleneck in scaling-up COVID-19 testing is the need for sophisticated instruments and well-trained healthcare professionals, which are already overwhelmed due to the pandemic. Moreover, the high-sensitive SARS-CoV-2 diagnostics are contingent on an RNA extraction step, which, in turn, is restricted by constraints in the supply chain. Here, we present CASSPIT (Cas13 Assisted Saliva-based & Smartphone Integrated Testing), which will allow direct use of saliva samples without the need for RNA extraction for SARS-CoV-2 detection. CASSPIT utilizes CRISPR-Cas13a based SARS-CoV-2 RNA detection, and lateral-flow assay (LFA) readout of the test results. The sample preparation workflow includes an optimized chemical treatment and heat inactivation method, which, when applied to 94 COVID-19 clinical samples, showed a 97% positive agreement with the RNA extraction method. With CASSPIT, LFA based visual limit of detection (LoD) for a given SARS-CoV-2 RNA spiked into the saliva samples was ~200 copies; image analysis-based quantification further improved the analytical sensitivity to ~100 copies. Upon validation of clinical sensitivity on RNA extraction-free saliva samples (n=76), a 98% agreement between the lateral-flow readout and RT-qPCR data was found. To enable user-friendly test results with provision for data storage and online consultation, we subsequently integrated lateral-flow strips with a smartphone application. We believe CASSPIT will eliminate our reliance on RT-qPCR by providing comparable sensitivity and will be a step toward establishing nucleic acid-based point-of-care (POC) testing for COVID-19. The diagnostic-challenge in SARS-CoV-2 detection is multifold, though some of these issues were addressed recently with the development of point-of-care (POC) testing, more innovative, specific, and sensitive methods are needed to provide accurate tests results and facilitate mass testing (Cheng et al., 2020) . One such challenge is the use of multi-step RNA extraction, a bottleneck that impedes mass testing for COVID-19. In this direction, RNA extraction-free assays are being developed and validated on clinical samples (Kriegova et al., 2020; Wee et al., 2020) . Initially, these methods were standardized on swab samples, and the results showed comparable sensitivity with that of RNA extraction methods. (Alcoba-Florez et al., 2020; Brown et al., 2020; Bruce et al., 2020; Grant et al., 2020; Hasan et al., 2020; Merindol et al., 2020; Srivatsan et al., 2020; Wee et al., 2020) . Although contradictory reports exist regarding sensitivity based on saliva than swab samples for SARS-CoV-2 Williams et al., 2020; Wyllie et al., 2020) , these samples are also directly used without RNA extraction (Vogels et al., 2020) . The direct use of these samples relies on RNA release with heat inactivation and a few chemical combinations. One such approach, previously validated in other viruses like Zika and Dengue, utilizes Hudson buffer (TCEP and EDTA) and heat-inactivation at 95 o C (Myhrvold et al., 2018) . Similarly, others have found that dual heat denaturation at 55 o C and 95 o C works better than a single heat inactivation step (Meyerson et al., 2020; Vogels et al., 2020) . Furthermore, a range of chemical reagents tested like Tris-EDTA, Tris borate, ionic and nonionic detergents, and RNA stabilizing reagents indicate the reliability of this workflow (Lalli et al., 2020; Ranoa et al., 2020) . While these methods have their limitations and advantages, all the studies unanimously suggest that RNA extraction-free SARS-CoV-2 detection works in saliva 5 based on Cas13a, which is validated in many biological samples including saliva, and can reliably detects bacterial and viral pathogens with both LFA and fluorescent-based readout (Gootenberg et al., 2017 Myhrvold et al., 2018) . The technique was named SHERLOCK (specific high-sensitivity enzymatic reporter unlocking), which has a single-base specificity and single-molecule sensitivity and precisely detects multiple pathogens in a single reaction (Gootenberg et al., 2017) . The SHERLOCK based diagnostics take advantage of high-end instrument free RPA or RT-RPA based amplification of the nucleic acids, which makes this approach amenable and straightforward for on-site and at-home testing while at the same time enhancing the sensitivity and specificity of the assay (Gootenberg et al., 2017 Myhrvold et al., 2018) . Recently, SHERLOCK has been standardized and validated for the detection of SARS-CoV-2 in clinical swab samples (Patchsung et al., 2020) . Similarly, tools like All-in-One Dual CRISPR-Cas12a (AIOD-CRISPR) or colorimetric LAMP assay using Cas12a were developed (Ding et al., 2020; Joung et al., 2020) . However, optimizing these methods utilized input RNA samples obtained using commercially available kits, which adds to the test's cost and testing time. As of now, we have not come across any study which has used SHERLOCK based detection on RNA extraction-free clinical saliva samples for COVID-19 diagnosis. We developed and clinically validated Cas13a integrated lateral-flow readout to detect SARS-CoV-2 in RNA extraction-free saliva samples. We have provided a simple workflow from sample processing to detecting the end-results using the visual paperbased test strip. Further, we have provided a quantitative estimate of the test strip signal readouts for the straightforward interpretation of the test results with a provision for field-deplorability and at-home testing. 6 We used plasmids containing S and N genes of SARS-CoV-2, respectively, to standardize RT-qPCR, using CDC-approved and in-house designed primers. Among a set of 8 primers tested, two primer pairs for S gene (SP-1, SP-4) and CDC verified primer for N gene (N1) generated a single amplicon, with S gene amplification slightly better than N1 at same plasmid DNA concentration. RT-qPCR further confirmed these results (Figure 1a we generated S gene synthetic fragments containing T7 polymerase corresponding to the region flanking amplified sequence by SP-1 primer. The synthetic DNA fragments were subsequently converted to RNA using an in-vitro transcription assay, following which the transcribed RNA was extracted, purified, and quantified (see Methods). We performed an RT-qPCR reaction using various purified RNA dilutions and plotted the corresponding Ct values against the known concentration ( Figure 1b ). Using SP-1 primer, we detected up to a single copy of RNA of S gene corresponding to Ct value <39.26. This LoD obtained for S gene agrees with analytical sensitivity defined for N1 gene (Vogels et al., 2020a) . Similarly, we used RNA extracted from four saliva samples of the volunteers to find any cross-reactivity. These samples were collected in 2018 for an unrelated study and stored at -80 o C. Similar amplification profile and LoD was obtained with spiked-in RNA in these samples, while lack of amplification without spiked-in RNA suggests that no-cross reactivity exists with the RNA obtained from the saliva samples ( Figure 1c ). After finding LoD, we used these same sets of primers and primer/probes for N1, RdRp, and E gene using a commercially available kit and performed RT-qPCR in the clinical saliva samples. 7 Initially, we obtained 102 clinical saliva samples from patients with matched swab samples tested at Safdarjung Hospital, New Delhi. Based on the swab results for which the hospital performed the RT-qPCR, 78 samples were positive, and 24 negative samples were negative. We extracted RNA from these samples using the viral RNA extraction kit (see Methods) followed by one step RT-qPCR analysis using the commercially available kit and compared the results with our in-house optimized protocol. Based on previous reports and the LoD derived for the S gene, we set the upper limit for detection at Ct < 40 to mark the sample as positive, and samples with Ct above 40 were marked negative. In the initial screening, 8 samples showed no detectable signal for internal control and hence were eliminated for the analysis. Among 72 saliva samples, we found 98.6% positive agreement with swab results with atleast one of the primer/probe sets tested. Individually, we found 95.8% agreement for N and S gene among individual genes, 94.4% for the E gene, and 93% for RdRp. Among the negative samples, all genes except N showed a 100% positive agreement, while one sample with a negative swab result showed a positive signal for the N gene. We excluded eight samples for the assay due to the undetectable internal control (IC) signal ( Figure 1d ). Comparative analysis of our in-house optimized protocol with a commercially available kit revealed a close correlation of S gene with N (r = 0.878) (Figure 1e3 , 1f3), which was comparatively better than comparison for S gene with E gene (Figure 1e1 , f1) and RdRp ( Figure 1e2 , f2). Thus, our in-house optimized RT-qPCR method is in high-agreement with the CDCapproved N-1 gene-based amplification. Further, these results confirm that saliva and swab samples have a high degree of consistency, and saliva can decisively detect SARS-CoV-2 in COVID-19 patients. Our results confirm the findings of the previously published reports that have demonstrated the use of saliva as a reliable clinical sample for detecting SARS-CoV-2 with All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 8 a detection limit and sensitivity in par with the nasopharyngeal and oropharyngeal swab (hereafter swab) (Fakheran et al., 2020; Procop et al., 2020) . While some other reports have shown slightly better analytical sensitivity of saliva samples (0.98 virus RNA copies/ml) than other biological fluids for SARS-CoV-2 detection (Wyllie et al., 2020), we did not perform a direct comparison. A major hurdle in COVID-19 testing is the need for viral RNA extraction, which poses a challenge to speed-up the testing. We envisioned to use a simple RNA extraction-free (hereafter, RNA_ExF) method and validate its analytical sensitivity on SARS-CoV-2 clinical samples. Recently, several RNA extraction-free methods were employed to test their analytical sensitivity in clinical samples; however, they have their own benefits and limitations (Alcoba-Florez et al., 2020; Brown et al., 2020; Bruce et al., 2020; Grant et al., 2020; Hasan et al., 2020; Merindol et al., 2020; Srivatsan et al., 2020; Wee et al., 2020) . Thus, we aimed to develop a more sensitive workflow for the RNA extraction-free testing of saliva samples. We initially optimized Proteinase K concentration under various heat inactivation conditions and tested by introducing 100 000 copies of the S gene into the normal saliva. We found that a concentration of 1.25mg and dual heat inactivation (37 degrees for 10 minutes and 95 o C for 5 min) exhibited better analytical sensitivity in comparison to heating of samples at 65 o C or with higher concentrations of Proteinase K (Figure 2a) . While, at all concentrations tested, the detection limit was relatively less than the samples spiked into RNAse free water. As saliva contains mucoproteins, which may interfere with the detection, we next tried mucoactive chemicals (sodium citrate and ammonium chloride) and mucolytic agent N acetylcysteine (NAC). Interestingly, we found that samples treated with NAC exhibited relatively better detection than other mucoactive agents. We also used 0.5% Triton X100 which exhibited no interference with detection and may further help release viral RNA when applied to clinical samples ( Figure 2a) . Thus, we found an optimal heat inactivation method (37 o C for 10 minutes and 95 o C for 5 minutes) and chemical composition which consists of proteinase K (1.25mg/ml), NAC (0.5%), and Triton X100 (0.5%). To obtain the LoD, we used spiked-in S standard RNA into the saliva samples and performed chemical treatment and heat inactivation. We could accurately detect as low as 10 copies of RNA/reaction, which though slightly lesser compared to a single copy detection when RNA was spiked-in water ( Figure 2b ). We further tested this buffer's sensitivity and heat inactivation on eight saliva samples collected from volunteers in 2018 for an unrelated study. In comparison to the RNA spiked in water, we could detect the amplification in all saliva samples, while a slight increase in Ct values was observed, with a difference in the mean Ct of 2.169 ± 0.9526 (Figure 2c1 , 2c2). Collectively, these results indicate that an optimized buffer and heat inactivation conditions are suitable for RNA detection in saliva samples. Based on the results that RNAlater solution maintains the RNA quality and detection sensitivity in clinical samples (above, Figure 1d ), we initially used RNAlater solution and guanidine hydrochloride (GuHCL) to collect the saliva samples for validation using RNA extraction-free workflow. We gave 3 sets of tubes for sample collection; the set-I with RNAlater, set-II with GuHCL, and set-III without any solvent. A total of 8 samples collected from the same patients were obtained and used for the analysis. We found that both GuHCL and RNAlater inhibited the assay detection with our optimized chemical and heat treatment. Simultaneously, direct use of the samples without preservatives showed better detection ( Figure 2d ). Thus, these All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; 10 results suggest that the collection of saliva directly into the collection tube without any chemical or RNA preservative is optimal for the RNA extraction-free detection. Next, we used the same optimized protocol to validate this workflow's sensitivity on 83 additional clinical samples. Each collected sample collected was divided into two separate tubes, one containing the RNAlater solution, and in the other tube, saliva was collected without any solvent. The samples collected in RNAlater were subjected to RNA extraction using the kitbased method, while samples collected without any solvent underwent chemical treatment and heat inactivation and were directly used for RT-qPCR analysis. Initially, we tested N and S primers' sensitivity with the RNA extraction-free method and performed the assay in four samples. A slightly better sensitivity was detected when we used N primer in samples with RNA extraction-free method ( Figure 2e ). This discrepancy in the two primers' sensitivity could be due to N primer's small amplicon size (72bp) than S (112bp). So, we used N primer and the RnaseP (RP) for the subsequent screening of the clinical samples. We used RP for initial screening to qualify the sample for comparative analysis between the RNA extraction and the extraction-free method. Only those samples with a detectable Ct value for RP (76 out of 79 samples) were qualified and further used for further analysis. As shown in Figure 2f , 76 samples of COVID-19 were tested with RNA extraction-based method, out of which 47 tested positive, while 29 showed no detectable signal and were marked negative. With our optimized RNA extraction-free workflow, 45 out of 47 samples showed positive and 29 out of 29 showed negative results, with an overall 95.7% agreement for positive test samples and 100% agreement for the negative samples (Figure 2f, g) . Thus, a high degree of correlation between RNA extraction-free workflow and RNA extraction method was obtained (r=0.98) (Figure 2h ). The two samples that showed no detectable signal in the RNA extraction-All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. 11 free method had a comparatively high Ct value obtained with the RNA extraction method (Ct: 37) (Figure 2i ). Though the Ct values were slightly higher with the RNA extraction-free method, with a difference between the means of the two methods at 2.545 ± 1.158, overall, these results suggest that saliva can be directly used for the detection of SARS-CoV-2 in clinical samples without the need for costly and time-consuming RNA extraction steps. This simple extraction free detection workflow overcomes the RNA extraction kits' supply chain limitation and maintains the analytical sensitivity in clinical samples. To determine the sample stability over time, we conducted the assay on SARS-CoV-2 saliva samples stored at room temperature. We observed no significant difference in the analytic sensitivity of the samples up to 6 hours of storage ( Figure 2j ). Thus, these results suggest feasibility of home collection of the saliva samples without technical assistance, cold storage, or viral transport medium. After validating RNA extraction-free detection of SARS-CoV-2 in saliva samples, we next explored detection methods, which are (1) relatively instrument-free, (2) previously validated for SARS-CoV-2, and (3) exhibit sensitivity and specificity at par or with RT-qPCR. To meet this criterion, we found the SHERLOCK-based detection method to be the most appropriate, which was also recently approved by the FDA. SHERLOCK relies upon the collateral activity of Cas13a to cleave the colorimetric or a fluorescent reporter once the target molecule is detected ( Figure 3a ). To use this method on our optimized workflow, we obtained commercially synthesized and previously verified crRNA sequences for the S gene spanning the region for which we validated the RT-qPCR assay (Supplementary Table 1) . Similarly, we also obtained All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; 12 crRNA corresponding to the Orf1ab gene (Zhang lab, MIT). Cas13a was isolated from an Addgene plasmid (#90097) which was a kind gift from the Zhang Lab, MIT. The plasmid was propagated and purified based on the published protocol (Patchsung et al., 2020) . Using the methodology employed by Kellner et al., we first performed SHERLOCK on standard RNA corresponding to S and Orf1ab to validate this method (Kellner et al., 2019) . Previous studies have shown that both fluorescence-based detection and lateral-flow readouts using paper-strip can be accurately used to demonstrate this method's working principle. Considering the ease-of-use, instrument-free detection, and cost-effectiveness, we choose the lateral-flow readout to validate this method and correlate it with the corresponding RT-qPCR data. A range of various dilutions of the standard RNA spiked-in control saliva was run, and the lateral-flow readouts labeled as positive or negative by visual detection. As shown in Figure 3b , a consistent increase in the positive signal in the test lane was obtained, which corresponded to the lowest RNA copy number of 200 copies/reaction for S gene with 100% detection sensitivity corresponding to Ct value 35.43. In comparison, the detection limit for Orf1ab was lower at 400 copies/reaction ( Figure 3c ). The LoD for S gene was slightly higher than the spiked RNA samples, as described previously (100 copies/reaction for S gene) (Joung et al., 2020) . This discrepancy is probably due to the sample processing and assay for detection. While the study by Zhen et al. used spiked in RNA samples followed by RNA extraction, and Cas12-based detection, we used RNA extraction-free samples and employed Cas13-based detection. Overall, these results confirm the SHERLOCK-based approach's validity and sensitivity integrated with visual lateral-flow readout. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; To make this tool affordable and accessible with a provision for field testing, we next performed SHERLOCK on the optimized RNA extraction-free saliva samples. We divided the samples into 4 groups based on the Ct values. Group I with Ct > 25; group II with Ct between 26-30; group III with Ct between 31-35, and group IV with Ct < 36, with 5 samples in each with RNA extraction-free saliva samples, we could obtain comparable sensitivity to the RNA extraction method for SARS-CoV-2 detection, as reported by others (Patchsung et al., 2020) . To further validate this approach's reliability, we performed a longitudinal detection analysis of the test samples. For this, we selected two patients repeatedly sampled until the time the test results were negative. In these patients, we found four samples collected at different time intervals after the symptom onset till the test results were negative. We used these samples and conducted RT-qPCR and SHERLOCK, followed by LFA. The lateral-flow readout accurately confirms the findings of the RT-qPCR results (Supplementary Figure 2b1, 2b2) . Taken together, these results confirm that SHERLOCK-based diagnosis provides a reliable high-end instrument free testing for SARS-CoV-2, without compromising the assay's analytical sensitivity. The limitation in LFA based detection for viral pathogens is that the signal obtained may exhibit false positives if minute RNAse contamination in the samples persists, which is highly concerning when using extraction free methods (which may non-specifically cleave the reporter molecule). A third test lane for detecting the non-specific activity of RNAse was introduced in All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; 14 the test strip to detect RNAse contamination (Patchsung et al., 2020) . However, it has largely remained unexplored to distinguish background signal from the actual signal in the test lane. Based on our results and previously published studies, SHERLOCK with visual readout performs exceedingly well with samples at a higher copy number of the analyte. Though, the interpretation of results become difficult when samples with very low copies of analytes are present, as a shallow background signal in the test lane is detected in some samples, which may be due to non-specific probe degradation, as stated previously (Patchsung et al., 2020) . Thus, to determine the background signal and differentiate between the noise and actual signal, we used ten replicates of negative control and samples with 50, 100, and 200 copies of RNA/reaction spiked-in saliva and subjected to chemical treatment and heat inactivation, respectively ( Figure 4b ). The LFA images were obtained and processed using Image J (see Methods). A threshold value was generated based on the ratio of signal intensity in test lane to control lane (T/C) corresponding to no RNA and three known concentrations of RNA. Based on the signal obtained from negative samples, we obtained a threshold value with a T/C ratio of 0.15, and below this ratio, the samples were labeled as negative ( Figure 4c ). By quantifying the signal of various dilutions, all samples with 200 copies of RNA showed T/C ratio above threshold (positive). In samples with 100 copies, nine showed positive T/C ratio, though most of these sample results were difficult to interpret by visual detection. Thus, we set this new LoD to 100 copies based on signal quantitation, which corresponded to a Ct value of about 36.32 ( Figure 4c ). Next, we quantified the signal from the respective positive and negative clinical samples. A 100% agreement was observed in the negative samples between visual and T/C ratio. Similarly, samples with Ct> 35 also showed 100% agreement. Surprisingly, among five samples with Ct< 36, two showed T/C signal above threshold, in contrast to one sample detected by the All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; 15 visual readout (Figure 4d ). Thus, while visual detection accurately detects the signal below a Ct value of 35, it suffers the detection limit in samples with very low viral RNA copy numbers. A semi-quantitative approach is more feasible for such samples, as quantitation is provided based on the ratio between the test and the control lane instead of solely relying on the signal at the test lane. Overall, these results suggest that LFA signals obtained using RNA extraction-free saliva samples can be precisely quantified and correlated with Ct values to give a fair estimate of the viral load (at higher Ct) and further enhance the analytical sensitivity. Presently, we are working on developing an automated mobile application which can be integrated with the LFA based detection to provide the user with quantitative end-results of the sample for field or at-home testing when instrument-free detection methods like SHERLOCK or DETECTR are used ( Figure 5 and Supplementary Video 1). Considering the impact of COVID-19 on global health and economy, it is imperative to have at our disposal field-deployable diagnostic methods that are robust, cost-effective, specific, and sensitive. The onset of COVID-19 pandemic has witnessed an upsurge in molecular diagnostics of SARS-CoV-2, but only a few of them have sensitivity and specificity comparable to the gold standard RT-qPCR testing, especially those based on nucleic acid detection such as CRISPR-Cas system (Broughton et al., 2020; Patchsung et al., 2020) . However, the major challenge to use these tools as a POC device for field testing or at-home tests persists. In addition, workflow of these methods is contingent on additional RNA extraction steps that require trained professionals and advanced instrumentation to perform the assay. Considering these fundamental challenges, we developed, optimized, and validated the use of a simple workflow to detect SARS-CoV-2 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 16 using saliva samples for the following reasons: (1) saliva can be self-collected with ease to minimize direct contact with the healthcare workers, and to minimize handling errors; (2) repeated sampling is feasible without incurring discomfort to the patient, with a reasonably uniform sample distribution; and (3) higher stability of the SARS-CoV-2 in saliva even when stored at room temperature (up to 7 days) (Vogels et al., 2020b) . Initially, we validated the analytical sensitivity of clinical saliva samples following RNAextraction and RT-qPCR assay for detection and found a close agreement with corresponding swab test results performed by the hospital. Next, we developed a simple workflow to detect SARS-CoV-2 in saliva samples without RNA-extraction. Optimization of this workflow was challenging as unlike other biological fluids, the molecular composition of saliva hinders RNA detection, besides being more amenable to RNases (Ochert et al., 1994; Ostheim et al., 2020) . Also, SARS-CoV-2 being an enveloped virus, the RNA release condition had to be optimized so that its degradation is minimized. Working on various chemical treatments and heat inactivation steps, we formulated a unique buffer composition containing the optimal concentration of Proteinase K, Triton X 100, and N-acetyl cysteine, followed by heat inactivation. Together this buffer and heat condition were sufficient to release SARS-CoV-2 RNA from the sample. On testing this workflow on clinical samples, we found a close agreement with the kit-based RNA extraction method, which is used for detection of SARS-CoV-2. Our results were thus consistent with previous reports on the use of RNA extraction-free detection of SARS-CoV-2 (Alcoba- Florez et al., 2020; Brown et al., 2020; Bruce et al., 2020; Grant et al., 2020; Hasan et al., 2020; Merindol et al., 2020; Srivatsan et al., 2020; Wee et al., 2020; Vogels et al., 2020b) . To further simplify the diagnostic workflow, we have optimized the RNA extraction-free detection of SARS-CoV-2 with two-(step)pot SHERLOCK, which relies on the collateral All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint activity of Cas13a -recently validated on clinical COVID-19 samples (Patchsung et al., 2020; Hou et al., 2020) . Our results indicate that the two-step(pot) SHERLOCK and RT-RPA reaction works exceedingly well in saliva samples with RNA extraction-free method, and found 98% positive agreement with the RT-qPCR data, with Ct >35, corresponding to about 200 copies/reaction. In view of the reports that most SARS-CoV-2 patients have a cut off Ct values around this range (33.5 for E; 33.5 for RdRp; and 34.5 for N gene) (Uhm et al., 2020) , our approach can be reliably used as an alternate to RT-qPCR with RNA extraction-free samples. The LoD which we obtained for extraction free saliva samples with a two-pot reaction is slightly lower than what others found when RNA-extraction methods were used (Patchsung et al., 2020) . This could be due to: (1) Ample amount of DNA, proteins, and other ions present in saliva, which interferes with the detection, irrespective of the type of detection method used, i.e., RT-qPCR or SHERLOCK (Ochert et al., 1994; Ostheim et al., 2020) . (2) preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 2020). Thus, there is scope to further improve the analytical sensitivity of RNA extraction-free workflow. Another disadvantage of paper-strip-based methods is that these test results are difficult to access for clinical studies. That includes data for survey, vaccine trials, or testing other therapeutic interventions. To overcome this challenge, we integrated the lateral-flow readout with a mobile application with provision for offline or online mode. The smartphone integrated workflow will thus provide an accurate estimate of the signal in test vs control lane, that will remove any ambiguity associated with visual detection. Further, this application will provide access to image files, patient clinical parameters, and LFA quantitative results. A video demonstration of a saliva-based workflow integrated with the mobile application is shown (Supplementary Video 1) . In summary, we provide a simple workflow for SARS-CoV-2 detection, which is a unique addition to the rapid, cost-effective, and straightforward diagnostic methods. Such userfriendly testing methods have an immediate application under the settings where trained professionals and costly instruments are limited. Further, owing to the high-sensitivity and highspecificity of Cas13a, our optimized workflow on saliva samples could provide a rapid and better alternative to the existing detection methods and speed up the testing. We envision that in coming time, CRISPR-Diagnostics based on either Cas13a or a similar method which utilizes Cas12 (Broughton et al., 2020) will be a better alternative to RT-qPCR based testing under resource limiting settings. Moreover, integrating CRISPR-Diagnostics with RNA extraction-free workflow and smartphone application will provide an alternative to error-prone rapid antigen tests to scale-up COVID-19 testing across resource-constrained areas and intensify trace, test, and treat strategy for COVID-19. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint The future of CRISPR-Diagnostics is promising as it provides rapid, accurate, low cost, and laboratory free genetic testing. Our study has shown feasibility of this diagnostic approach with RNA extraction-free saliva samples which can be extended to home testing. There is still scope to further improve these tools, especially for use in fully instrument free settings and without pre-amplification step. While a recent study has provided an improved version of CRISPR-Diagnostics, which bypasses the amplification step (Fozouni et al., 2020) , more optimization is needed to provide a fully instrument-free diagnostic platform. The work was indented to develop a simple workflow for SARS-CoV-2 detection. The clinical samples collected for the work were used after obtaining Institutional ethical clearance from the Safdarjung Hospital (IEC/VMMC/SJH/Project/2020-07/CC-06). Patient consent was obtained to collect the samples according to the ICMR GCP guidelines. A total of 185 clinical saliva samples were collected from Safdarjung hospital from June till October 2020, New Delhi. The saliva samples were collected from the patients at the same time when swab was collected for COVID-19 testing by the hospital. Initially all the samples were collected in RNAlater solution for validation of the saliva-based detection of SARS-Cov-2 with RT-qPCR. Subsequently, samples were divided into 2 parts; (part 1) was collected in RNAlater solution, and (part 2) was collected in tubes containing proteinase K (1.25mg), Triton X 100 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 20 (0.5%), and NAC (0.5%). All the samples were processed in NABL certified (MC-3486) and ICMR approved Diagnostic laboratory for COVID-19 testing following the regulatory guidelines and protocol (360 Diagnostic and Health Services, Noida, U.P., India). Samples were collected at different time interval in the hospital and stored at -20 o C until carried to the facility for further processing. The time between sample collection to processing was around 3-5 days. Plasmids corresponding to S and N genes were received as a gift from Krogan laboratory, Department of Cellular and Molecular Pharmacology (San Francisco, CA 94158, USA). These same plasmids are now available with Addgene (#141382 for S gene and #141391 for N gene). After propagating, plasmid DNA was isolated using commercially available DNA isolation kit (Vivantis technologies). For RT-qPCR, 1ng of the DNA was used for respective genes and a set of 8 primers were optimized. These primers were synthesised in-house using online primer design tools or obtained based on previously validated sequences like N-1 primers. The list of the primers is given in the Supplementary Table 1. Synthetic gene fragments for S gene and Orf1ab corresponding to the sequences given in the Supplementary Table 1, were obtained from Xcelris Genomics. The gene fragments were synthesised along with the T7 promoter sequence. Various heat-inactivation steps, chemical components and buffers were used to find the optimal assay condition for the detection of the viral RNA. Heat inactivation optimization was done at All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 21 37 o C for 15 mins and 95 o C for 5 mins. Non-ionic detergents like Triton X 100 (Sigma) and Proteinase K (Sigma, vivantas, and promega) were used at various combinations to find the optimal reaction composition. Optimum detection of spiked-in S gene was obtained at a concentration of 1.25 mg/ml for proteinase K and 0.5% of triton X100 with either two step heatinactivation (65 o C for 5 mins or 10 mins and 95 o C for 5 mins) as well as with a single heatinactivation step (RT for 15 mins and 95 o C for 5 mins). Further, in order to minimize the interference of the mucoprotein in saliva, mucoactive agents like sodium citrate (Sigma) and ammonium chloride (Sigma); and mucolytic agent N-acetyl cysteine (Sigma) were used at various concentrations. RNA extraction was performed as recommended (Qiagen viral RNA extraction kit). 140ul sample was processed according to the protocol as per manufacturer's instruction. Final elution of the RNA was done in 30ul of the elution buffer. 2 ul of the extracted RNA/reaction was used in one-step RT-qPCR analysis using the commercially available RT-qPCR kit for SARS-CoV-2 which contains there targeting genes E, N and RdRp along with the internal control RNase P. Similarly, we also used our in-house optimized primers/probes for validation. RT-qPCR was performed on Rotor gene Q (Qiagen) with the recommended reaction condition for the commercial kit (Allplex, Seegene). For SP-1 and SP-4, the following RT-qPCR conditions were used: Initial denaturation 95ºC for 5 minutes, second cycle of the reaction include denaturation at 95ºC for 30 sec, annealing at 62ºC for 30 sec, extension at 72ºC for 30 sec for 40 cycles; All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint To determine the limit of detection (LoD), we generated synthetic gene fragments for S and Orf1ab corresponding to the regions for which crRNA has been previously validated (Zhang Lab protocol, MIT and Supplementary Table 1 ). Both single stranded gene fragments were obtained commercially (Xceliris Genomics). Each (1ng/µl) oligonucleotide fragments were first converted into double strand by end point PCR then purified (0.5 µg/µl) of the double stranded DNA fragment was used as a template for invitro transcription reaction using in vitro transcription kit (T7 Ribomax, promega, cat no-P1320). The invitro transcribed RNA oligonucleotide fragment was then purified using RNA cleanup kit (vivantis, cat no-GF TR 050) and eluted in a final volume of 50 ul. The purified RNA of the respective gene fragments was used as the standard to determine LoD. Various dilutions of S gene standard RNA were made in nuclease free water corresponding to 10 0 to 10 6 copies/ul. Similarly, the standard RNA (10 6 copies/ul) was spiked into the saliva samples for optimizing various chemical and heat-inactivation conditions. For the expression of Cas13a protein, BL-21 cells were transformed by heat shock method with pC013 -Twinstrep-SUMO-huLwCas13a plasmid. The transformed cells were grown overnight in 10 ml LB media containing 100 µg/ml ampicillin antibiotic at 37ºC/180 rpm in incubator shaker. Next day, 5 ml of the overnight grown culture was inoculated in 1liter LB media containing 100µg/ml Ampicillin. After reaching the growth of the culture to OD between 0.4-0.6, the culture was incubated at 4ºC for 30 mins. Before induction of the protein, 1ml of culture was taken for SDS-PAGE analysis. Expression of the protein was induced by adding 1 ml/0.5 M IPTG and 2% glycerol to pre-chilled culture and incubated in cooling (21ºC) incubator shaker for All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi. org/10.1101 org/10. /2020 23 16 hours at 300rpm. After that, the culture was harvested by centrifugation at 6000 rpm for 10 minutes at 4ºC. supernatant was discarded and pellet was lysed in the lysis buffer. The cell pellet was resuspended in lysis buffer (20 mM Tris-HCl pH-8.0, 500 mM NaCl, 1 mM DTT, 1x protease inhibitor cocktail sigma, and 0.5 mg/ml lysozyme). The cell resuspension was lysed by sonication (Sartorius Stedim) using 50% pulse amplitude (on 10s and off 20s) until completely lysed. The lysate was centrifuged at 12,000 rpm for 30 mins at 4 °C. The protein was then applied to a HiTrap SP HP column equilibrated with equilibration buffer (20 mM Tris-HCl pH 8.0, 1 mM DTT, 500 mM NaCl, 15mM imidazole). The supernatant fraction was passed five times from the column for complete binding of the protein. After washing with the binding buffer to remove nonspecific binders, the recombinant His6-SUMO-LwaCas13a was eluted in a linear gradient (with increasing the concentration of the imidazole from 20-500mM) of elution buffer. The best elution was obtained at 100 mM imidazole. Elution was done in the volume of 5 ml. For the cleavage of His tag, the eluted fraction was supplemented with 20µl of sumoprotease (invitrogen #125880-18,1U/µl) and 7.5 µl of NP-40. The reaction mixture was added to the column and incubated at 4°C for overnight with gentle shaking. Next day, cleaved native LwaCas13a protein was obtained by draining the column. After draining of the column, it was washed with elution buffer containing 500mM imidazole to ensure the complete cleavage of Histag from LwaCas13a protein. The drained native protein was subjected to concentration to 0.25ml using centrifugal spin filter (50 MWCO) at 4000 rpm for 15 minutes at 4°C. Then 5ml of protein storage buffer (1M Tris-HCl pH-7.5, 5M NaCl, 5% glycerol and 10 µl DTT) was added to the same filter and again All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 24 centrifuged at the same condition. Storage buffer containing native Lw-Cas13a protein was diluted to 2mg/ml in storage buffer and stored at -20°C, as described previously (Kellner et al., 2020) . As described previously (Gootenberg et al., 2017; Kellner et al., 2019) , RT-RPA was performed using the commercially available RPA kit (TwistDx). First, each RPA tube was divided into four reaction tubes by diluting the lyophilized mix in the RPA buffer (40 ul). For LoD, forward and reverse RT-RPA primers were added 1ul (10 µM stock) to each reaction tube along with 1 µl of reverse transcriptase enzyme (EpiScript). 4ul of various dilutions of the standard RNA for S and Orf1ab were used as template (0 to 4000 copies/reaction). For RNA extraction-free samples, a total of 8 ul of sample input was used and the reaction components were adjusted accordingly. The reaction was initiated by adding 0.7 µl of magnesium acetate (280 mM stock). All the reagents were prepared and mixed at 4 o C. Finally, the reaction mix was incubated at 42 o C for 25 minutes and tapped in between after every 3-5 minutes. Particularly, tapping of the samples was found to exhibit better results than without and is highly recommended henceforth. SHERLOCK assay of the above reaction mix was performed in a separate 1.5ml Eppendorf tube which contains 1 µl of the Cas13a either from extracted pool or commercially obtained source (at 63.3 µg/ml concentration), 1 µl RNase inhibitor (20 U per µl stock; invitrogen), 0.6 µl T7 RNA polymerase (50 U per µl stock; Lucigen), 1µl of crRNA for the respective genes (Synthego), 1µl of MgCl2 (120mM), 0.8µl of rNTP mix (100mM), 2ul of cleavage Buffer (400mM Tris pH 7.4), and 1µl (20uM) reporter. 6 µl of the RPA mix from RNA All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint 25 extracted samples were used per reaction. The final volume of the reaction mix was adjusted to 20 µl with RNAse free water. The reaction mix was incubated at 37 °C for 25 mins. The SHERLOCK reaction mix was subjected to lateral-flow assay using the commercially available test trips and buffer (Millenia Biotec). To the above reaction mix 80 µl buffer (HybriDetect assay) was added, provided with the kit. The visual readout of the test results was obtained by dipping the test strips ((Milenia Biotech1T) into the respective 1.5ml Eppendorf tubes. To provide a semi-quantitative analysis of the lateral-flow readout, we used Fiji image J software to analyze the signal in the respective T and C bands. The corresponding band intensity of the test lane (T) and control lane (C) were calculated using integrated density parameter. The image of each strip was captured using a mobile phone (width=50, height=220). For the quantification, the image was further cropped to image size 40by220 which thus removed the outliers from the image. Any background noise from the image was subtracted using the rolling ball background subtraction method by keeping radius=50. All the images(40by220) were further thresholded by applying lower threshold value 0 and the upper threshold values between 240-245. Finally, the single band was segmented from each image in the frame size of 30by30 and integrated density was analyzed for the respective bands. The threshold value of T/C was calculated and found to be 0.15 above which the samples could be labelled as positive. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint The mobile application for detection and quantification of lateral-flow strips was developed using machine learning tools. The algorithm was implemented using the OpenCV package v.4.3.0 in Python 3.7.3. The Android App was developed with Android Studio v4.2 RC 2 (Google) with Java 8 and Gradle v4.1.0. To provide a clean user interface, the main screen was limited to a "COVID-19 Test" bottom tab button that opens up an in-app camera view to capture the image followed by custom image edit options. The image acquisition is only allowed through the mobile application to accurate documentation of taken images and test results. The image captured in the application is obtained as a Uri object, that is used for conversion to a byte's array. In order to obtain the image analysis outcome, the bytes array obtained is passed as an argument to the Python backend script running through Chaquopy v6.3.0 that is a software development kit used in Android development environment. The bytes array image is further processed by the integrated image analysis module of the application. In the first image preprocessing step, the acquired image is converted to grayscale and region of interest is localized in the image through adaptive thresholding and edge detection techniques. In the next step, the perspective and orientation corrections are performed on the image which are present due to uncontrolled mobile imaging. The shape of the strip and marker are two primary cues used for the aforementioned corrections. The resultant image is centered around the localize region of interest and cropped to obtained the strip area image to a fixed scale. The sample and control band image regions are detected from the cropped image based on the known strip structure and darker intensity profile of the bands. The mean intensity of each band region is obtained and ratio of the sample to control band in then calculated and displayed on the mobile application to the user along with the cropped strip image. The mean intensity of the sample band region is scaled to the All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 27 range of 0-1 w.r.t a pre-defined reference point in situation when control band has weak visual appearance (signature). Statistical analysis of the samples was performed using GraphPad8 Prism. Pearson correlation coefficient was used for correlation analysis. Non-parametric t test was used to compare the mean difference between two data sets as mentioned in the results. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 34 noninvasive specimen for detection of sars-cov-2. J. Clin. Microbiol. doi:10.1128 /JCM.00776-20. Wyllie, A. L., Fournier, J., Casanovas-Massana, A., Campbell, M., Tokuyama, M., Vijayakumar, P., et al. (2020 . Saliva or Nasopharyngeal Swab Specimens for Detection of SARS-CoV-2. N. Engl. J. Med. doi:10.1056/nejmc2016359. Xia, S., and Chen, X. (2020) . Single-copy sensitive, field-deployable, and simultaneous dual- All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 30 40 S N S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16 S17 S18 S19 S20 Saliva SP-1 SP-1 Ct Value 23 27 27 25 24 24 25 24 23 29 28 29 29 30 29 29 28 29 31 30 31 31 31 31 30 30 22 37 37 22 19 19 19 19 18 19 31 26 32 20 32 31 21 34 23 34 24 26 30 29 28 26 29 29 30 27 34 33 28 29 33 33 32 30 31 31 33 33 32 32 36 31 30 24 26 21 22 21 19 19 19 32 28 37 25 36 33 24 36 26 38 27 22 23 24 24 24 24 23 23 25 27 26 25 25 26 25 24 29 24 24 23 20 24 22 22 20 22 22 30 25 25 27 22 21 19 25 24 22 24 18 21 27 25 21 30 36 22 18 20 21 20 37 25 20 25 19 21 19 24 20 20 26 21 21 22 21 22 21 22 20 22 20 19 21 26 24 26 A B C 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. ; https://doi.org/10.1101/2020.11.07.20227082 doi: medRxiv preprint Orf1ab was found to be higher than S gene at 400 copies/reaction. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this this version posted November 10, 2020. Visual LFA + + + + + + + + + + + + + + + + ---------Visual LFA + + + + + + + + + + + + -+ -+ -+ ---------- To obtain semi-quantitative analysis of the data, 10 images of paperstrips corresponding to 200, 100, 50, and 0 RNA copies as shown in Figure 3b and Figure 4b , were subjected to image quantitation. The threshold value (T/C ratio) was obtained based on the signal in the control (C) and test (T) lane. The T/C ratio was considered positive above the background value of 0.15. Based on T/C ratio, the LoD was found to be 100 copies of RNA per reaction. (d) Visual results of paper-strips shown in Figure 4a , were subjected to signal quantitation and T/C ratio was calculated. One sample (blue arrow head, ID: 39), which was difficult to characterize by visual LFA detection, was correctly characterized as positive using T/C ratio. Fast SARS-CoV-2 detection by RT-qPCR in preheated nasopharyngeal swab samples CRISPR-Cas12-based detection of SARS-CoV-2 Validation of an extraction-free RT-PCR protocol for detection of SARS-CoV2RNA Direct RT-qPCR detection of SARS-CoV-2 RNA from patient nasopharyngeal swabs without an RNA extraction step CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity. Science (80-. ) Diagnostic Testing for Severe Acute Respiratory Syndrome-Related Coronavirus 2: A Narrative Review Ultrasensitive and visual detection of SARS-CoV-2 using all-in-one dual CRISPR-Cas12a assay. Nat. 30 point-of-care testing for SARS-CoV-2 in a community screening setting shows low sensitivity Saliva as a diagnostic specimen for detection of SARS-CoV-2 in suspected patients: A scoping review Direct detection of SARS-CoV-2 using CRISPR-Cas13a and a mobile phone Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a and Csm6 Nucleic acid detection with CRISPR-Cas13a/C2c2 Extraction-free COVID-19 (SARSCoV-2) diagnosis by RT-PCR to increase capacity for national testing programmes during a pandemic Detection of SARS-CoV-2 RNA by direct RT-qPCR on nasopharyngeal specimens without extraction of viral RNA Development and evaluation of a rapid CRISPR-based diagnostic for COVID-19 Point-of-care 31 testing for COVID-19 using SHERLOCK diagnostics Clinical performance of a semi-quantitative assay for SARS-CoV2 IgG and SARS-CoV2 IgM antibodies SHERLOCK: nucleic acid detection with CRISPR nucleases CRISPR-Cas guides the future of genetic engineering Direct-RT-qPCR Detection of SARS-CoV-2 without RNA Extraction as Part of a COVID-19 Testing Strategy: From Sample to Result in One Hour Rapid and extraction-free detection of SARS-CoV-2 from saliva with colorimetric CRISPR/Cas Systems towards Next-Generation Biosensing SARS-CoV-2 detection by direct rRT-PCR without RNA extraction A community-deployable SARS-CoV-2 screening test using raw saliva with Field-deployable viral diagnostics using CRISPR-Cas13. Science Inhibitory effect of salivary fluids on PCR: Potency and removal Overcoming challenges in human saliva gene expression measurements Clinical validation of a Cas13-based assay for the detection of SARS-CoV-2 RNA DNA detection using recombination proteins A Direct Comparison of Enhanced Saliva to Nasopharyngeal Swab for the 33 Preliminary support for a "dry swab, extraction free" protocol for SARS-CoV-2 testing via RT-qPCR Development and Evaluation of a Novel Loop-Mediated Isothermal Amplification Method for Rapid Detection of Severe Acute Respiratory Syndrome Coronavirus Consistent Detection of 2019 Novel Coronavirus in Saliva Patterns of viral clearance in the natural course of asymptomatic COVID-19: Comparison with symptomatic non-severe COVID-19 Analytical sensitivity and efficiency comparisons of SARS-CoV-2 RT-qPCR primer-probe sets SalivaDirect: A simplified and flexible platform to enhance SARS-CoV-2 testing capacity Rapid direct nucleic acid amplification test without rna extraction for sars-cov-2 using a portable pcr thermocycler Dr. Ahmad is thankful for UGC for the start-up grant support (F.4-5/2018 FRP Start-up grant). Authors declare that there is no conflict of interest. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this this version posted November 10, 2020. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.