Development of Web Application for the Comparison of Segment Variability with Sequence Evolution and Immunogenic Properties for Highly Variable Proteins: An Application to Viruses

Abstract

AbstractHigh rate of mutation and structural flexibilities in viral proteins quickly make them resistant to the host immune system and existing antiviral strategies. For most of the pathogenic viruses, the key survival strategies lie in their ability to evolve rapidly through mutations that affects the protein structure and function. Along with the experimental research related to antiviral development, computational data mining also plays an important role in deciphering the molecular and genomic signatures of the viral adaptability. Uncovering conserved regions in viral proteins with diverse chemical and biological properties is an important area of research for developing antiviral therapeutics, though assigning those regions is not a trivial work. Advancement in protein structural information databases and repositories, made by experimental research accelerated the in-silico mining of the data to generate more integrative information. Despite of the huge effort on correlating the protein structural information with its sequence, it is still a challenge to defeat the high mutability and adaptability of the viral genomics structure. In this current study, the authors have developed a user-friendly web application interface that will allow users to study and visualize protein segment variabilities in viral proteins and may help to find antiviral strategies. The present work of web application development allows thorough mining of the surface properties and variabilities of viral proteins which in combination with immunogenicity and evolutionary properties make the visualization robust. In combination with previous research on 20-Dimensional Euclidian Geometry based sequence variability characterization algorithm, four other parameters has been considered for this platform: [1] predicted solvent accessibility information, [2] B-Cell epitopic potential, [3] T-Cell epitopic potential and [4] coevolving region of the viral protein. Uniqueness of this study lies in the fact that a protein sequence stretch is being characterized rather than single residue-based information, which helps to compare properties of protein segments with variability. In current work, as an example, beside presenting the web application platform, five proteins of SARS-CoV2 was presented with keeping focus on protein-S. Current web-application database contains 29 proteins from 7 viruses including a GitHub repository of the raw data used in this study. The web application is up and running in the following address: http://www.protsegvar.com.