New Computational Approach for Peptide Vaccine Design Against SARS-COV-2

Abstract

The design for vaccines using in silico analysis of genomic data of different viruses has taken many different paths, but lack of any precise computational approach has constrained them to alignment methods and some alignment-free techniques. In this work, a precise computational approach has been established wherein two new mathematical parameters have been suggested to identify the highly conserved and surface-exposed regions which are spread over a large region of the surface protein of the virus so that one can determine possible peptide vaccine candidates from those regions. The first parameter, w, is the sum of the normalized values of the measure of surface accessibility and the normalized measure of conservativeness, and the second parameter is the area of a triangle formed by a mathematical model named 2D Polygon Representation. This method has been, therefore, used to determine possible vaccine targets against SARS-CoV-2 by considering its surfacesituated spike glycoprotein. The results of this model have been verified by a parallel analysis using the older approach of manually estimating the graphs describing the variation of conservativeness and surface-exposure across the protein sequence. Furthermore, the working of the method has been tested by applying it to find out peptide vaccine candidates for Zika and Hendra viruses respectively. A satisfactory consistency of the model results with pre-established results for both the test cases shows that this in silico alignment-free analysis proposed by the model is suitable not only to determine vaccine targets against SARS-CoV-2 but also ready to extend against other viruses.

Combatting future variants of SARS-CoV-2 using an in-silico peptide vaccine approach by targeting the spike protein

“…The main focus of this in silico experiment will be to test whether our computational approach [19] is able to recommend the most suitable peptide vaccine targets against SARS-CoV-2 such that they evade all possible mutations in the frequently observed strains of concern of the virus and are not concealed by any neighbouring amino acids. The factors involved in evaluating this hypothesis include studying the mutation probability and solvent accessibility across the spike protein sequence and checking the epitope potential and autoimmunity of the selected peptide targets.…”

“…A ranking list for all the 12-length peptides is prepared in descending order of their w values. The top-ranked peptides are then “grouped” as per their position in the sequence, by following certain necessary conditions [19] . This eventually gave 16 grouped “peptide zones”.…”

“…This step deals with calculating a score for these zones with the help of a mathematical model, namely, “2D Polygon Representation” [19] . The model uses the normalized versions of three different parameters – average ASA, average 1/PV, and the span of each zone.…”

“…The epitope potential of the regions was primarily judged by studying the T-cell epitopes given by the MHC-II binding predictions in Immune Epitope Database Analysis Resource (IEDB-AR) [21] . Finally, we checked if the qualified zones pose any significant chance of causing autoimmune threats [19] , using NCBI Basic Local Alignment Search Tool (BLAST) server [22] .…”

“…In this article, we have described a novel computational approach that takes into account the surface-exposed protein of SARS-CoV-2, that is, spike glycoprotein, and recommends the best possible vaccine candidates for the virus [19] . The method offers two primary advantages in comparison with other competent techniques.…”