The new coronavirus SARS-CoV-2 that emerged late 2019 is still causing a pandemic of severe acute respiratory syndrome or COVID-19. The pandemic affected over 500 million people and claimed more than six million lives (Johns-Hopkins-University, 2022), thus posing a difficult challenge to the scientific and healthcare communities throughout the world (Malik et al., 2020; Tutelyan et al., 2020; World-Economic-Forum, 2020). Indeed, SARS-CoV-2 is a positive RNA virus that is constantly evolving through the accumulation of various type of mutations (Zhao et al., 2020; Banoun, 2021; Majumdar and Niyogi, 2021; Singh et al., 2021). Even though the majority of these mutations do not affect the virus infectious properties and have no real impact on the progress of the pandemic (Ashoor et al., 2021), some may enhance specific viral attributes that give the virus a selective advantage (Gobeil et al., 2021). Any new variant endowed with selective advantage(s) would favor the virus persistence and nurture the pandemic. Therefore, watching and predicting how the pandemic evolves and communicate it to the public is of paramount importance. The global surveillance of the pandemic is based on multidisciplinary approaches including epidemiological, genetic, structural, and clinical data (Agency, U.H.S, 2021; Campbell et al., 2021; England, P.H, 2021; France, S.P, 2021). This involves the use of a set of relevant criteria to categorize the variants. Toward this end, all international and national health and sanitary authorities have set various strategies to control the evolution of the SARS-CoV-2 pandemic. On 31 May 2021, the World Health Organization (WHO) announced a labeling system to categorize the variants into different levels of priority to better organize the global monitoring and research, and ultimately organize the “infodemic” and communicate more effectively with the public about the adequate response to the emergence of new variants of SARS-CoV-2 (https://www.who.int/). WHO has first developed a system to facilitate naming SARS-CoV-2 variants in addition to the existing nomenclature systems for naming and tracking SARS-CoV-2 genetic lineages established by GISAID (https://www.gisaid.org/), Nextstrain (https://nextstrain.org/), and Pango (https://cov-lineages.org/). These nomenclatures are mostly used by the scientific research community. For practical reasons (particularly to ease the communication), WHO has settled to name the SARS-CoV-2 emerging variants using the Greek letters alphabet sequence (α, β, γ, δ…). Since the SARS-CoV-2 virus is showing high genetic variability (Toyoshima et al., 2020; Yazdani et al., 2021; Dubey et al., 2022), WHO has established a labeling system for the variants into variant of concern (VOC), variant of interest (VOI), and variant under monitoring (VUM). This labeling system is based on definitions related to variant phenotypic attributes such as transmissibility, disease presentation, effect on current diagnostic tests, and response to available vaccines. This system was set to prompt and harmonize the actions needed to control the spread of a given variant. While the variants that emerged sequentially were named α, β, γ, and δ, the last emerging one was named omicron. Of interest is that the United States Centers for Disease Control and Prevention (CDC) has also adopted this system of variant classification and labeling but added more labels and labeling criteria and different labels’ change policy (https://www.cdc.gov/). Indeed, while keeping the labels, VOC, VOI, and VUM (calling the later “VBM” for Variant Being Monitored). In addition, the CDC uses an extra label which is “Variant Of High Consequences” (VOHC). As a result, some of the currently known variants are given different labels by each agency. Indeed, according to the CDC, there is no SARS-CoV-2 variant labeled VOI as of December 1, 2021. Furthermore, variants α, β, and γ that are currently labeled VOC by WHO have been deescalated to the VBM label by the CDC as of September 2021 (https://www.cdc.gov/). These discrepancies reflect different views on the labeling of SARS-CoV-2 variants and consequently the use of the labels to set public health actions. On the other hand, in both systems, variant labels can change with more data accumulating for a particular variant.

In this work, we have undertaken an evaluation of the system developed by WHO and adapted by the CDC to label the SARS-CoV-2 variants. We carried out a review of the classification criteria and analyzed how WHO and the CDC use these criteria to label the SARS-CoV-2 variants. Then, we used a comparative computational predictive approach to study the S protein mutations that characterize the five SARS-CoV-2 VOCs. We concluded that computational predictions provide a good ground of evidence for a rapid and more accurate labeling system.

Despite significant global healthcare measures and social mitigation efforts along with the availability of a number of vaccines, the SARS-CoV-2, COVID-19 pandemic entered its third year and the numbers of cases are soaring worldwide (https://www.gisaid.org/, https://coronavirus.jhu.edu/map.html). This is mainly due to the emergence of new and more fit viral variants that continue to fuel this pandemic. To create awareness about any additional health issue a new variant may cause, WHO, the world’s most influential health agency, has developed a labeling system to classify the new variants, in addition to the conventional nomenclature. The CDC has adopted this labeling system but introduced major changes. Among the labels used by WHO and the CDC is the label “Variant of Concern or VOC”. The word “concern” is synonymous of anguish, anxiety, and apprehension. It naturally enjoins fear and disturbance, which predispose taking immediate action and making odd changes. Indeed, when a new variant is labeled VOC, countries are inclined to close their borders and take drastic countermeasures. This was particularly striking when a number of countries banned travels from South Africa, which was the first country to describe and report the genomic sequence of the omicron variant. While preventive and cautionary actions are mandatory to control pandemics, these should be based on solid scientific evidence. However, it takes time and coordinated efforts for the scientific community to generate the data needed to accurately label new variants according to criteria such as transmissibility, disease severity, and change in the epidemiological pattern, immune escape, and resistance to previously neutralizing antibodies, efficacy of existing therapies, or vaccine efficiency. Meanwhile, the rapid pandemic progress requires timely response including a good communication system. In this study, we highlight some flaws in the labeling system developed by WHO mainly to ease communication with the national health authorities and the public. Indeed, in addition to the non-appropriate wording for labeling emergent variants, the combination of criteria used to define a variant label as shown in WHO and the CDC web sites, is not accurate. Indeed, in their definitions of the different labels, both agencies build a combination of criteria using the prefix “OR” but not “AND”. This introduces a confusion that is amplified by the CDC use of more labels, and different formulation and different combinations of the criteria used to attribute a label to a new variant. According to WHO, a label is supposedly assigned to a variant through a comparative assessment with the previous ones. Even though SARS-CoV-2 variants are primarily detected upon the virus genomic sequence changes and particularly mutations in important functional regions, the genetic variations do not clearly appear among the viral attributes WHO and the CDC use to formulate their definition of the different variants. The labeling system developed by WHO and the definitions of the criteria for each label used by both WHO and the CDC do not mention a comparative assessment of the genetic variations but rather retain clinical, epidemiological transmissibility and other non-genetic viral attributes. Such parameters need a substantial time to be accurately determined. It seems that the current labeling system relies more on fear from emergent genetic variations, “the super killer virus”, than evidence about the impacts of mutations (Caini et al., 2018). Thus, it seems premature and not appropriate to give a new variant the VOC label. Even when a greater transmissibility can be established, this does not necessarily mean greater severity. For example, the very transmissible H1N1 influenza virus variant was not as severe as many other influenza viruses and the epidemic has naturally faded away (Grubaugh et al., 2020). With reference to the WHO variant labeling system, H1N1 could have been labeled VOC while it has never been one. Therefore, to solve the dilemma between the necessity of rapidly labeling a new variant and the lack of scientific evidence, a good approach is to thoroughly analyze the genetic variations observed to rapidly generate the best data on the potential attributes of the new viral variant. To achieve this analysis, the computational prediction of variant effect on protein stability, function, and interaction is a very useful way to determine the variant “importance”. Several methods can be used based on the availability or non-availability of the 3D structure of the protein. In this study, we show that computational prediction and in silico comparative studies between new variants and older more characterized ones can provide good insights into the impact of the mutations and thus the potential behavior of the emergent variants. Indeed, the observation that most of the mutations in SARS-CoV-2 VOCs map to the solvent-exposed regions along with the comparison of antigenicity predictions (Crooke et al., 2020; Vashi et al., 2020; Chen et al., 2021) represents a good basis to make a fair assumption on a potential immune escape and/or a likely reduction of antibody neutralization and to anticipate an evaluation of vaccine effectiveness. In addition, the comparative analysis of the S protein 3D structure between the five SARS-CoV-2 VOCs showed an identical overall folding as demonstrated by the RMSD values and Tm scores of each model. Meanwhile, the comparison of contact residue maps gave a fine-tuning of the structural divergence, which suggests potential functional differences that need to be further analyzed.

Furthermore, computational studies of the mutation effect on the thermodynamic stability of the S protein/ACE2 complex and the comparative analysis of the pattern of polar interactions of the different variant with the ACE2 receptor give good indications to predict transmissibility and virulence and draw some plausible epidemiological scenario. Indeed, the effect of mutations can be computed as single or combined mutations. The data we obtained with the combined mutations of the omicron variant that have a much higher number of mutations than the other variants show that this variant engages into a more stable interaction with ACE2 than the β and γ variants do and has more interactions than the delta variant.

For the other variant labeling criteria, several studies discussed the importance of the Glu484Lys mutation in the interaction with ACE2 and immune escape (Greaney et al., 2021b; Makdasi et al., 2021). This was also observed with the beta and gamma variant that carry Glu484Lys mutation. It was reported that the beta variant has reduced antibody neutralization compared to the delta variant. Moreover, beta’s resistance to neutralizing antibodies increased by 9.4-fold to convalescent plasma and 10.3- to 12.4-fold for sera from individuals who have been vaccinated (Liu et al., 2021a). In addition, it was suggested that new variants with the same mutation might bear new challenges for current vaccines or monoclonal antibody therapies (Krause et al., 2021; Zhou and Wang, 2021).

Meanwhile, our analysis of the pattern of polar interactions between the different variants shows that omicron has an extra 13 hydrogen bonds as compared to variant delta. In addition, we noticed the presence in the more transmissible alpha and delta variants of an extra salt bridge between the S protein Glu 484 and ACE2 lysine 31 and the presence of an extra salt bridge between the S protein Gln 493 and ACE2 Glu 35 in the omicron variant. This observation, combined with the observed high number of H bonds and the thermodynamic stability data, predicts a potentially more efficient entry into host cells and enhanced transmissibility of the omicron variant, which has been confirmed since the description of the omicron mutation profile. Nevertheless, relying on the study of a single mutation or the computing of one biophysical feature of a variant structure to predict how a viral attribute would evolve is not sufficient. Furthermore, multiple genes can control epidemiologically relevant viral attributes such as the mode of transmission and virulence. Therefore, it is recommended to integrate the data on multiple mutations with computation of various viral structural features to make the best predictions and attribute the right label to a given variant.

In conclusion, the system of SARS-CoV-2 labeling developed by WHO and amended by the CDC has some major flaws. Relying on the integrated biophysical and structural data generated from computational comparative predictions of the likely behavior of a new variant would help in the rapid and accurate labeling of emergent variants. Meanwhile, given our provisional and incomplete knowledge and the uncertain nature of the COVID-19 pandemic, it would be wise to operate in epistemic self-abnegation, use the best tools and knowledge we have at hand, and introduce revisions whenever new evidence becomes available.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

DA: in silico analysis, methodology, data curation, writing, and editing. MM: mutations’ review, illustrations, figures, and tables. KT: data retrieval from data banks and formatting. SC: analysis and illustration of the WHO and CDC variant-labeling criteria. NA: data retrieval and iconography. NK: review of the in silico study. AR: data cross-checking and coordination. M-DF: project conception, work design, data analysis, writing, editing, and supervision. All authors contributed to the article and approved the submitted version.

Grant LS_COVID 19, 2020 from the Arabian Gulf University.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.