Researchers have developed EVScape, a predictive tool that identifies which SARS-CoV-2 strains are most likely to evade host immunity. Utilizing a deep generative model and structural information, the tool offers early warning about viral variants, aiding in the timely development of effective vaccines and treatments.
By Neha MathurOct 12 2023Reviewed by Susha Cheriyedath, M.Sc. In a recent article published in the journal Nature, researchers developed EVEscape, a tool to forecast which severe acute respiratory syndrome coronavirus 2 strains have the highest potential to escape host immunity.
EVScape working mechanism EVscape used a method called EVE20 to predict the effects of substitution mutations in viral proteins on its fitness, i.e., sustained ability to replicate within the host and cause infection. The second EVScape component, called antibody accessibility, identified potential antibody binding sites. It computed the accessibility of each amino acid residue, i.e., the likelihood of antibodies recognizing a mutated viral protein, using its negatively weighted amino acid residue-contacts across three-dimensional structures accounting for protrusion from the core and conformational flexibility.
Related StoriesWithin the whole S, EVEscape scores were biased towards the receptor-binding motif of the receptor-binding domain and the neutralizing supersite in the N-terminal domain , implying it identified the most immunogenic domains of viral antigens without prior knowledge of any antibody epitopes.
This model captured the most frequent indels at site 144 and in the top decile of random and pandemic indel predictions. They retrained EVE models with the addition of 11 million novel sequences, and even this model captured epistatic shifts between Wuhan and BA.2 strains.