key: cord-0968761-5crm4kkh
authors: Perez-Romero, Carmina A.; Tonda, Alberto; Mendoza-Maldonado, Lucero; Coz, Etienne; Tabeling, Patrick; Vanhomwegen, Jessica; Claassen, Eric; Garssen, Johan; Kraneveld, Aletta D.; Lopez-Rincon, Alejandro
title: Design of Specific Primer Sets for the Detection of SARS-CoV-2 Variants of Concern B.1.1.7, B.1.351, P.1, B.1.617.2 using Artificial Intelligence
date: 2021-10-15
journal: bioRxiv
DOI: 10.1101/2021.01.20.427043
sha: a40e4fda24dda4736a8d932e0098ea414770c024
doc_id: 968761
cord_uid: 5crm4kkh

As the COVID-19 pandemic continues, new SARS-CoV-2 variants with potentially dangerous features have been identified by the scientific community. Variant B.1.1.7 lineage clade GR from Global Initiative on Sharing All Influenza Data (GISAID) was first detected in the UK, and it appears to possess an increased transmissibility. At the same time, South African authorities reported variant B.1.351, that shares several mutations with B.1.1.7, and might also present high transmissibility. Earlier this year, a variant labelled P.1 with 17 non-synonymous mutations was detected in Brazil. Recently the World Health Organization has raised concern for the variants B.1.617.2 mainly detected in India but now exported worldwide. It is paramount to rapidly develop specific molecular tests to uniquely identify new variants. Using a completely automated pipeline built around deep learning and evolutionary algorithms techniques, we designed primer sets specific to variants B.1.1.7, B.1.351, P.1 and respectively. Starting from sequences openly available in the GISAID repository, our pipeline was able to deliver the primer sets for each variant. In-silico tests show that the sequences in the primer sets present high accuracy and are based on 2 mutations or more. In addition, we present an analysis of key mutations for SARS-CoV-2 variants. Finally, we tested the designed primers for B.1.1.7 using RT-PCR. The presented methodology can be exploited to swiftly obtain primer sets for each new variant, that can later be a part of a multiplexed approach for the initial diagnosis of COVID-19 patients.

SARS-CoV-2 mutates with an average evolutionary rate of 10 −4 nucleotide substitutions per site each year 1 . As the pandemic of SARS-CoV-2 continues to affects the globe, researchers and public health officials constantly monitor the virus for variants of concern (VOC) with acquired mutations that may pose a treat to global health such as: a higher rate of transmissibility, change in epidemiology, virulence, clinical presentation, mortality, vaccine/therapeutics resistance, or decrease in effectiveness of public health measures 2 .

On December 14th, 2020, Public Health authorities in England reported a new SARS-CoV-2 variant, [3] [4] [5] , which belongs to the B.1.1.7 (which include all Q. lineages used for fine geographical localization of the variant) Pango lineage 6, 7 , GRY clade from GISAID (Global Initiative on Sharing All Influenza Data) 3, 8, 9 , Nextstrain clade 20I(V1) 10 . This was the first VOC raised by the World Health Organization (WHO), and was recently termed as the Alpha variant 2 . This variant presents 14 nonsynonymous mutations, 6 synonymous mutations, and 3 deletions. The multiple mutations present in the viral RNA encoding the WHO are those on 417N position 2 . The L452R mutation has been shown to increase SARS-CoV-2 viral infectiousness and replication 65 . Interestingly. L452R and E484Q have been found to disrupt the interfacial interactions of the Spike RBD with neutralizing antibodies 66, 67 . Furthermore, the L452R has been found to be a positive adaptive mutation driving the spread of similar variants with these mutations such like the ones in California 68 . The T478K is located at the interface of the Spike/ACE2 interaction domain 15, 69 , and enhance stabilization of the Spike RBD with the ACE2 complex 70 . The Delta variant has been show to lead to a 64% higher household transmission rate 71 , higher viral loads 72, 73 and increase in hospital and ICU admissions when compared with the Alpha variant 74, 75 . It has been associated to an increase in infections and excess deaths across the globe 64, 72, 76, 77 . The Delta variant has been shown to be resistant to monoclonal antibody therapy treatment with bamlanivimab 25, 28, 47, 78 , but remains susceptible to casirivimab, imdevimab and etesevimab 25, 28, 47, 50 . Furthermore, the Delta variant has reduce neutralization by convalescent 25 and modestly reduced sensitivity to Comirnaty/BNT162b2 mRNA vaccine 75, [78] [79] [80] and ChadOx-1 vaccine sera 36, 75, 80 requiring two-doses for full protection. Although several cases of reinfection from the Delta variant have been documented in partially and fully vaccinated individuals and previously Covid-19 infected patients 81 , the severity of infection and risk of hospitalization remains lower in vaccinated individuals 74, 82 . Therefore, the main concern remains in the greater burden seeing so far on health-care services due to outbreaks on unvaccinated population 74 . The Delta variant has spread like wildfire across the globe rapidly since its explosive rise in India on May, it has now become the major VOC globally rapidly displacing other variants 8, 35, 71, 73 . Given its high prevalence and clinical implications many countries have started implemented stronger vaccine policies 83 , considering the need of additional doses for vulnerable population 84 and the need for vaccine updates to combat waning immunity 85 .

Given the rapid spread of these VOC, and their impact on global health, it has become of utmost importance for countries to be able to rapidly identify and detect them. Several diagnostic kits have been proposed and developed to diagnose SARS-CoV-2 infections. Most kits rely on the amplification of one or several genes of SARS-CoV-2 by real-time reverse transcriptasepolymerase chain reaction (RT-PCR) 86, 87 . Recently, Public Health England was able to identify the Alpha variant through their national surveillance system which allowed them to notice the rise in Covid-19 positive cases in South Easter England. The Alpha variant was detected through the increase in S-gene target failure (negative results) from the otherwise positive target genes (N, ORF1ab) in the three target gene assay in their RT-PCR diagnostic tests, and random whole genome sequencing of some of this positive samples 3, 67 . In the case of the Beta 38, 88, 89 and Gamma 53, 90 variants were identify through the increase in positive Covid-19 cases and deaths and by random whole genome sequencing of positive Covid-19 samples, through their national and international diagnosis and surveillance system efforts. In the case of the Delta variant, most countries rely on the clinical differences between the variants and if the RT-PCR has a S fall out (indicating Alpha variant) or not indicating a plausible Delta variant 75 . Although several PCR methods and kits for the identification of this variants have been proposed 91, 92 , they usually rely on the well know Spike protein mutations, which often are shared with other VOIs and variants of SARS-CoV-2, and rapidly get obsolete due to the virus evolution 93 . Making sequencing the only reliable way to diagnose it, however this is cost prohibiting and many times not available in lower income countries.

In a previous work 94 , we developed a methodology based on deep learning, able to generate a primer set specific to SARS-CoV-2 in an almost fully automated way. Then, we reduced the necessary time by half using evolutionary algorithms 95 . When compared to other primers sets suggested by GISAID, our approach proved to deliver competitive accuracy and specificity. Our results for SARS-CoV-2 detection, both in-silico and with patients, yielded 100% specificity, and sensitivity similar to widely-used diagnostic qPCR methods. One of the main advantages of the proposed methodology was its ease of adaptation to different viruses or mutations, given a sufficiently large number of complete viral RNA sequences. In this work, we improved the existing semi-automated methodology, making the pipeline completely automated, and created primer sets specific for the SARS-CoV-2 variants B.1.1.7, B.1.351, P.1, B.1.6173.*, and B.1.1.519 in under 10 hours for each case study. The developed primer sets, tested in-silico, proved not only to be specific, but also to be able to distinguish between the different variants. In addition, we have validated our B.1.1.7 primers by RT-PCR, finding them to be specific to B.1.1.7 and sensitive enough for detection by this method. With this new result, we believe that our method represents a rapid and effective diagnostic tool, able to support medical experts both during the current pandemic, as new variants of SARS-CoV-2 may emerge, and possibly during future ones, as still unknown virus strains might surface.

Variant B.1.1.7 As explained in the Methods section, the first step is to run a Convolution Neural Network (CNN) classifier on the data. This yields a classification accuracy of 99.66%. Secondly, from an analysis of the features constructed by the CNN, we extract 7,127 features, corresponding to 21-bps sequences. Next, we run a state-of-the-art stochastic feature selection algorithm 10 times, to uncover the most meaningful '21-bps' features for the identification of variant B.1.1.7. While the best result corresponds to a set of 16 '21-bps' features, using only one is enough to obtain over 99% accuracy.

These features represent good candidates for forward primers. From the 10 runs, we get 10 different '21-bps' features: 5 out of the 10 point to mutation Q27stop (C27972T), 3 point to mutation I2230T (T6954C), and 2 to a synonymous mutation (T16176C). Using Primer3Plus we compute a primer set for each of the 10 features, using sequence EPI_ISL_601443 as the reference sequence. Only the two '21-bps' features that include mutation T16176C are suitable for a forward primer. The two features are ACCTCAAGGTATTGGGAACCT and CACCTCAAGGTATTGGGAACC: it is easy to notice that the two features are actually part of the same sequence, just displaced by a bps, and therefore generate the same reverse primer CATCACAACCTGGAGCATTG.

For further analysis, we check the presence of the signature mutations of B.1.1.7; T1001I, A1708D, I2230T, SGF 3675-3677 deletion, HV 69-70 deletion, Y144 deletion, N501Y, A570D, P681H, T716I, S982A, D1118H, Q27stop, R52I, Y73C, D3L and S235F of variant B.1.1.7 96 . To verify the presence of mutations, we generate 21-bps sequences, with 10 bps before and after the mutation, and search for their presence: e.g., mutation N501Y (A23063T) corresponds to sequence CCAACCCACT T ATGGTGTTGG.

Using the generated '21-bps' sequences for the mutations, we can also test them as forward primers using Primer3Plus, which yields TGATATCCTTGCACGTCTTGA in spike gene S982A as the only possible forward primer candidate, this sequence can be used for multiplex testing (Fig. 1) . Again, the first step is to run a CNN classifier on the data. This yields a classification accuracy of 99.88%. Next, we translate the CNN weights into 6,069 features, each representing a 21-bps sequence. we run the feature selection algorithm 10 times, which results in only one meaningful feature achieving over 99% accuracy. The 10 runs point to the same mutation, K417N, with 6 different sequences but only one acceptable as a primer candidate. The resulting sequence is CTCCAGGGCAAACTGGAAATA, with a reverse primer TGCTACCGGCCTGATAGATT.

We check whether 21-bps sequences containing the signature mutations T265I, K1655N, K3353R, SGF 3675-3677 deletion, L18F, D80A, D215G, R246I, K417N, E484K, N501Y, A701V, 242-244 del, Q57H, S171L, P71L, and T205I of variant B.1.351 96 could function as a forward primer. From the Primer3Plus results using EPI_ISL_678597, mutations D215G (TTAGTGCGTGGTCTCCCTCAG) and Q57H (CTGTTTTTCATAGCGCTTCCA) can be used as forwards primers (Fig. 2) . 

The CNN classifier on the data yields a classification accuracy of 100%, in the 28 samples. Next, we translate the CNN weights into 727 features, each representing a 21-bps sequence. We run the feature selection algorithm 10 times, which results in only one meaningful feature achieving 100% accuracy. The 10 runs point to the same sequence, around the synonymous mutation T733C ACTGATCCTTATGAAGACTTT. The sequence cannot be used as a primer, as the Tm is too low. To compensate, we displaced the sequence two bps to the left and added a bps to rise the Tm. This procedure gave us sequence GGCACTGATCCTTATGAAGACT, of size 22 bps, with reverse primer TTCGGACAAAGTGCATGAAG.

We check whether the sequences containing signature mutations S1188L, K1795Q, E5665D, SGF 3675-3677 deletion, T20N/L18F, P26S, D138Y, R190S, K417T, EA84K, N501Y, H655Y, T1027I, E92K, ins28269-28273 AACA, and P80R of variant P.1 53 could function as a forward primers. We put mutation T20N and L18F into the same '21-bps' sequence, given their proximity. From the Primer3Plus results using EPI_ISL_792683 , 3 of the generated sequences using mutations can be forward primers candidates; K417T (CAAACTGGAACGATTGCTGAT), H665Y (AGGGGCTGAATATGTCAACAA) and P80R (AATAGCAGTCGAGATGACCAA) (Fig. 3 ). 

A single-nucleotide mutation may not be enough to work as a specific primer for detecting SARS-CoV-2 variants. Thus, from the analysis of the characteristics mutations for each variant and the results of our pipeline, we created a list of primer sets based upon 2 or more mutation for each variant. For variant B.1.1.7 we created 3 primer sets, for variants B.1.351, and P.1 we created 2 options for each one, and for B.1.617.2 we created 3 different combinations (Table 1 ). Using Primer3Plus for in-silico simulations, we tried to maintain acceptable levels of temperature and an 18-22 bps size. Nevertheless, it was not always possible e.g. for P.1 variant the forward primers contain sequences size 25 bps. Another thing to consider is the product size, depending of which primer is going to be used the size will vary, as each primer falls into at least one mutation. The frequency of appearance for the different sequences is in Fig. 6 .

A wide variety of diagnostic tests have been used by high-throughput national testing systems around the world, to monitor the SARS-CoV-2 infection 86 . The arising prevalence of new SARS-CoV-2 variants such as B.1.1.7, B.1.351, P.1 and B.1.617.2 have become of great concern, as most RT-PCR tests to date are not be able to distinguish these new variants, not being designed for such a purpose. Therefore, public health officials most rely on their current testing systems and whole viral RNA sequencing results to draw conclusions on the prevalence of new variants in their territories. An example of such case has been seen in the UK, where the increase of the B.1.1.7 SARS-CoV-2 variant infection in their population was identified only through an increase in the S-gene target failure in their three target gene assay (N+, ORF1ab+, S-), coupled with sequencing of the virus and RT-PCR amplicons products 3 . Researchers believe that the S-gene target failure occurs due to the failure of one of the RT-PCR probes to bind, as a result of the 69-70 deletion in the SARS-CoV-2 spike protein, present in B.1.1.7 3 . This 69-70 deletion, which affects its N-terminal domain, has been occurring in several different SARS-CoV-2 variants around the world 10, 99 and has been associated with other spike protein receptor binding domain changes 9 . Due to the likeliness of mutations in the S-gene, assays relying solely on its detection are not recommended, and a multiplex approach is required 86, 87, 100 . This is consistent with other existing primer designs like CoV2R-3 in the S-gene 101 , that will also yield negative results for the B.1.1.7 variant, as the reverse primer sequence is in the region of mutation P681H. A more in-depth analysis of S-dropout positive results can be found in Kidd et al. 102 .

Given the concern for the increase in prevalence of the new variants SARS-CoV2 B. In this way, health authorities could better evaluate the medical outcomes of this patients, and adapt or inform new policies that could help curve the rise of variants of interest. Although the rest of the proposed primer sets delivered by our automated methodology will still require laboratory testing to be validated, our methodology can enable the timely, rapid, and low-cost operations needed for the design of new primer sets to accurately diagnose new emerging SARS-CoV-2 variants and other infectious diseases. In a first step, we train a convolution neural network (CNN) using the training and testing sequences obtained from GISAID. The architecture of the network is shown in Fig 7, and is the same as the one previously reported in 94 . Next, if the classification accuracy of the CNN is satisfying (> 99%), using a subset of the training sequences we translate the CNN weights into '21-bps' features, necessary to differentiate between the targeted variant samples and all the others. The length of the features is set as 21 bps, as a normal length for primers to be used in PCR tests is usually 18-22 bps. Then, we apply recursive ensemble feature selection (REFS) 104, 105 to obtain a reduced set of the most meaningful features that separate the two classes. Finally, we simulate the outcome of treating the most promising features obtained in the previous step as primers, using Primer3Plus 103 in the canonical sequences EPI_ISL_601443 for B.1.1. 

Another approach to generating primers is to use EAs. This approach has the advantage that we can parallelize the procedure and reduce the time required to get the regions of interest in comparison to the CNN-based method. In comparison to the 16 hours required to the CNN approach, each single run lasts around 62 minutes with 5 threads on a 64-bit Windows 10 laptop with Intel Xeon E-2186M. For the EA, We create a set of individuals of size 21-bps randomly considering the available samples of the variant of interest (e.g. B.1.1.519) and other variants. Next, we calculate what is known as cost function, which is given by the following:

with w p , w c , w n , w t representing the weights associated to each term.

P(I) is evaluating the presence of the sequence selected as candidate primer I inside training samples labeled with the variant of interest, and its absence from samples of other variants, T is the number of samples in the training set, s i is the i-th sample in the training set. Function P is defined as:

where L(s) returns the class label of sample s. In other words, P(I, s i ) equals 1 if sequence I is found inside a sample with the same class label as sample s k , the origin of sequence I. So, if the 21-bps sequence I is found inside a sample that does not belong to the variant of interest, or is not found in a sample that belongs to the variant of interest, the solution is penalized.

The second term of the weighted sum takes into account the GC content of the candidate primer:

where I(i) represents the base in position i inside sequence I. The following element of the weighted sum is N , defined as:

that takes into account the presence of N symbols in the sequence, indicating an error in the read. The ideal primer candidate should only contain A, C, G, or T values.

The final term tackles the requirement of having a melting temperature T m centered around 60 • . Specialized literature 103 provides an equation to compute T m for a sequence I: T m (I) = 81.5 + 16.6 * log 10 ([Na+]) + 41 * C (I) − 600/l(I) (6) where C (I) is the content of C and G bases in sequence I, as described in Equation 4 , [Na+] is the molar sodium concentration, and l(I) is the length of sequence I, in bps. We used the value of [Na+] = 0.2 as described in 103 , while l(I) = 21 by design. The term taking into account T m will then be:

The EA is set with a population of size µ = 200, generating offspring of size λ = 200. The entire population is replaced by its offspring at each generation, using a (λ , µ) replacement strategy, with a tournament selection of size τ = 2, a mutation acting on integer values, a one-point crossover, and a stop condition set on 100 generations. For more details, please refer to 95 .

Experimental evaluation of B.1.

Amplification efficiency of the designed primer sets were evaluated using viral RNA extracts from two sequenced SARS-CoV-2 strains: the original Wuhan strain 210207 (GISAID N°EPI_ISL_437689) and VOC B.1.1.7 strain (GISAID N°E PI_ISL_683466). Viral RNA were extracted from infected cell culture supernatants using the NucleoSpin Dx Virus kit (Macherey-Nagel), following the manufacturers' protocol. Viral RNA extracts (5 µL) were analyzed either using the IP2/IP4 dualplex real-time reverse-transcriptase (RT)-PCR assay, developed by following the Pasteur Institute and targeting conserved regions of the SARS-CoV-2 RdRP gene 98 , or primer set B.1.1.7-1 (Table 1) , using the LightCycler EvoScript RNA SYBR Green I Master kit (Roche). Both RT-PCR assays were conducted on a LightCycler® 480 System (Roche), using the thermal cycling program described in the Pasteur Institute protocol 98 . From the GISAID repository we downloaded 10,712 SARS-CoV-2 sequences on December 23, 2020. After removing repeated sequences, we obtained a total of 2,104 sequences labeled as B.1.1.7, and 6,819 sequences from other variants, for a total of 8,923 samples. B.1.1.7 variant samples were assigned label 1, and the rest were assigned label 0 for the CNN discovery described in detail in 94 . Next, from the found combinations and known mutations we generated primer sets and test them in 2,096,390 SARS-CoV-2 sequences downloaded in August 11 th , 2021, where 1,051,740 sequences are B.1.1.7. The total number of sequences by lineage is in the Table 2 on the supplemenatary material. From the GISAID repository, we downloaded 28 non-repeated sequences of the P.1 variant on January 19, 2021. We added the 28 sequences to the 8,323 sequences of several other variants, including B.1.1.7 and B.1.351, for a total of 8,351 sequences. We assigned label 1 to sequences belonging to variant P.1, and 0 to the rest of the samples using CNN to find the primers. Next, we generated primer sets and test them in 2,096,390 SARS-CoV-2 sequences downloaded in August 11 th , 2021 where 59,692 sequences are P.1. From the GISAID repository, we downloaded 836 sequences of the B.1.617.2 variant on May 5th, 2021. We added sequences to 6,819 sequences of other variants, and we assigned label 1 to sequences belonging to variant B.1.617.2, and 0 to the rest of the samples using EAs to find the primers. Finally, we generated primer sets and test them in 2,096,390 SARS-CoV-2 sequences downloaded in August 11 th , 2021 where 366,831 sequences are B.1.617.2.

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

Tracking sars-cov-2 variants

Investigation of novel SARS-COV-2 variant: Variant of Concern

Rapid increase of a SARS-CoV-2 variant with multiple spike protein mutations observed in the united kingdom

Assignment of epidemiological lineages in an emerging pandemic using the pangolin tool

A dynamic nomenclature proposal for sars-cov-2 lineages to assist genomic epidemiology

Global initiative on sharing all influenza data-from vision to reality

Preliminary genomic characterisation of an emergent sars-cov-2 lineage in the uk defined by a novel set of spike mutations

Nextstrain: real-time tracking of pathogen evolution

Deep mutational scanning of SARS-CoV-2 receptor binding domain reveals constraints on folding and ACE2 binding

Adaptation of SARS-CoV-2 in BALB/c mice for testing vaccine efficacy

The furin cleavage site of sars-cov-2 spike protein is a key determinant for transmission due to enhanced replication in airway cells

A multibasic cleavage site in the spike protein of sars-cov-2 is essential for infection of human lung cells

A tale of three sars-cov-2 variants with independently acquired p681h mutations in new york state

Functional evaluation of proteolytic activation for the sars-cov-2 variant b. 1.1. 7: role of the p681h mutation

Increased transmission of sars-cov-2 lineage b. 1.1. 7 (voc 2020212/01) is not accounted for by a replicative advantage in primary airway cells or antibody escape

Spike mutation d614g alters sars-cov-2 fitness

Sars-cov-2 spike-protein d614g mutation increases virion spike density and infectivity

Estimated transmissibility and severity of novel SARS-CoV-2 Variant of Concern 202012/01 in England. Available Github (2020). Online; accessed 26

Estimated transmissibility and impact of sars-cov-2 lineage b. 1.1. 7 in england

The sars-cov-2 b. 1.1. 7 variant and increased clinical severity-the jury is out

Increased mortality in community-tested cases of sars-cov-2 lineage b. 1.1. 7

Sars-cov-2 variant b. 1.1. 7 is susceptible to neutralizing antibodies elicited by ancestral spike vaccines

Reduced sensitivity of sars-cov-2 variant delta to antibody neutralization

Bamlanivimab does not neutralize two sars-cov-2 variants carrying e484k in vitro

Antibody resistance of sars-cov-2 variants b. 1.351 and b. 1.1. 7

Current status of therapeutic monoclonal antibodies against sars-cov-2

Infection and mrna-1273 vaccine antibodies neutralize sars-cov-2 uk variant

Sars-cov-2 b. 1.1. 7 sensitivity to mrna vaccine-elicited, convalescent and monoclonal antibodies

mrna-1273 vaccine induces neutralizing antibodies against spike mutants from global sars-cov-2 variants

Efficacy of chadox1 ncov-19 (azd1222) vaccine against sars-cov-2 variant of concern 202012/01 (b. 1.1. 7): an exploratory analysis of a randomised controlled trial

Emergence of SARS-CoV-2 b.1.1.7 lineage -united states

Tracking the international spread of sars-cov-2 lineages b. 1.1. 7 and b. 1.351/501y-v2

Rapid displacement of sars-cov-2 variant b. 1.1. 7 by b. 1.617. 2 and p. 1 in the united states

Sars-cov-2 b. 1.617. 2 delta variant emergence and vaccine breakthrough

New ay lineages and an update to ay.4-ay

Emergence and rapid spread of a new severe acute respiratory syndrome-related coronavirus 2 (sars-cov-2) lineage with multiple spike mutations in south africa

Enhanced receptor binding of sars-cov-2 through networks of hydrogen-bonding and hydrophobic interactions

Structure of the sars-cov-2 spike receptor-binding domain bound to the ace2 receptor

Identification of sars-cov-2 spike mutations that attenuate monoclonal and serum antibody neutralization

Complete mapping of mutations to the sars-cov-2 spike receptor-binding domain that escape antibody recognition

Comprehensive mapping of mutations to the sars-cov-2 receptor-binding domain that affect recognition by polyclonal human serum antibodies

Sensitivity of SARS-CoV-2 B.1.1.7 to mrna vaccine-elicited antibodies

Escape from neutralizing antibodies by sars-cov-2 spike protein variants

Molecular dynamic simulation reveals e484k mutation enhances spike rbd-ace2 affinity and the combination of e484k, k417n and n501y mutations (501y.v2 variant) induces conformational change greater than n501y mutant alone, potentially resulting in an esca

Fact sheet for health care providers emergency use authorization (eua) of bamlanivimab and etesevimab

501y. v2 and 501y. v3 variants of sars-cov-2 lose binding to bamlanivimab in vitro

Safety and efficacy of the chadox1 ncov-19 (azd1222) covid-19 vaccine against the b. 1.351 variant in south africa

Fact sheet for health care providers emergency use authorization (eua) of regen-covtm (casirivimab and imdevimab

Estimates of severity and transmissibility of novel south africa sars-cov-2 variant 501y

Genomic characterisation of an emergent sars-cov-2 lineage in manaus: preliminary findings

The emergence of novel sars-cov-2 variant p. 1 in amazonas (brazil) was temporally associated with a change in the age and sex profile of covid-19 mortality: A population based ecological study. The Lancet Reg

Sudden rise in covid-19 case fatality among young and middle-aged adults in the south of brazil after identification of the novel b. 1.1. 28.1 (p. 1) sars-cov-2 strain: analysis of data from the state of parana

Increased resistance of sars-cov-2 variant p. 1 to antibody neutralization

Genomic characterisation of an emergent sars-cov-2 lineage in manaus: preliminary findings

Resurgence of covid-19 in manaus, brazil, despite high seroprevalence

Reinfection by the sars-cov-2 p. 1 variant in blood donors in manaus

mrna vaccine-elicited antibodies to sars-cov-2 and circulating variants

Covid-19 weekly epidemiological update: Data as received by who from national authorities, as of

Covariants: SARS-CoV-2 mutations and variants of interest

Sars-cov-2 variants of concern and variants under investigation in england

for Disease Prevention, E. C. & Control. Emergence of SARS-CoV-2 B.1.617 variants in india and situation in the eu/eea

An emerging sars-cov-2 mutant evading cellular immunity and increasing viral infectivity

Convergent evolution of sars-cov-2 spike mutations, l452r, e484q and p681r, in the second wave of covid-19 in maharashtra

Possible link between higher transmissibility of B. 1.617 and B.1.1.7 variants of SARS-CoV-2 and increased structural stability of its spike protein and hace2 affinity

Acquisition of the l452r mutation in the ace2-binding interface of spike protein triggers recent massive expansion of sars-cov-2 variants

Preliminary report on sars-cov-2 spike mutation t478k

Sars-cov-2 spike mutations, l452r, t478k, e484q and p681r, in the second wave of covid-19 in maharashtra, india

Increased household transmission of covid-19 cases associated with sars-cov-2 variant of concern b. 1.617. 2: a national casecontrol study

Sars-cov-2 variant delta rapidly displaced variant alpha in the united states and led to higher viral loads

Infection with the sars-cov-2 delta variant is associated with higher infectious virus loads compared to the alpha variant in both unvaccinated and vaccinated individuals

Hospital admission and emergency care attendance risk for sars-cov-2 delta (b. 1.617. 2) compared with alpha (b. 1.1. 7) variants of concern: a cohort study

Sars-cov-2 delta voc in scotland: demographics, risk of hospital admission, and vaccine effectiveness

Sars-cov-2 variants of concern and variants under investigation in england

of Health Metrics, I. & (IHME), E. Covid-19 results briefing

Sars-cov-2 variant b. 1.617 is resistant to bamlanivimab and evades antibodies induced by infection and vaccination

Sars-cov-2 b.1.617 emergence and sensitivity to vaccine-elicited antibodies

Effectiveness of covid-19 vaccines against the b. 1.617. 2 (delta) variant

Severe sars-cov-2 breakthrough reinfection with delta variant after recovery from breakthrough infection by alpha variant in a fully vaccinated health worker. Front

Effectiveness of covid-19 vaccines against the b.1.617.2 (delta) variant

Good reasons to vaccinate: mandatory or payment for risk?

Effectiveness of pfizer-biontech and moderna vaccines in preventing sars-cov-2 infection among nursing home residents before and during widespread circulation of the sars-cov-2 b. 1.617. 2 (delta) variant-national healthcare safety network

How to redesign covid vaccines so they protect against variants

Molecular diagnostic technologies for covid-19: Limitations and challenges

Early transmission of sars-cov-2 in south africa: An epidemiological and phylogenetic report

Major new lineages of sars-cov-2 emerge and spread in south africa during lockdown

Genomic characterization of a novel sars-cov-2 lineage from rio de janeiro

Mutation-specific sars-cov-2 pcr screen: Rapid and accurate detection of variants of concern and the identification of a newly emerging variant with spike l452r mutation

Rapid base-specific calling of sars-cov-2 variants of concern using combined rt-pcr melting curve screening and sirph technology

Limitation of screening of different variants of sars-cov-2 by rt-pcr

Classification and specific primer design for accurate detection of sars-cov-2 using deep learning

Design of specific primer sets for sars-cov-2 variants using evolutionary algorithms

Detection of sars-cov-2 variants in switzerland by genomic analysis of wastewater samples

Sars-cov-2 variants of concern and variants under investigation in england

Protocol: real-time rt-pcr assays for the detection of sars-cov-2, institut pasteur, paris. World Heal

Recurrent independent emergence and transmission of sars-cov-2 spike amino acid h69/v70 deletions

Mutations on covid-19 diagnostic targets

The progression of sars coronavirus 2 (sars-cov2): Mutation in the receptor binding domain of spike gene

S-variant sars-cov-2 is associated with significantly higher viral loads in samples tested by thermofisher taqpath rt-qpcr

Primer3plus, an enhanced web interface to primer3

Automatic discovery of 100-mirna signature for cancer classification using ensemble feature selection

Machine learning-based ensemble recursive feature selection of circulating mirnas for cancer tumor classification

ALR and AT made the programming, data collection, and experiments in silico. EC, ADK and JG made the experiment and study design. EC and PT made the laboratory testing. CAP and ALR wrote the the article

The authors declare no competing interests.

Supplementary Material