Designing of a Novel Candidate Multi-epitope Vaccine to boost Immune Responses against SARS‐COV‐2 using Immunoinformatics and Machine Learning based Approach

Abstract

Background: The fast development of an effective vaccine is the major demand for protection against the SARS-COV-2 virus outbreak. With the vast volume of data and the requirement for automatic abstract property learning, machine learning (ML) as a branch of artificial intelligence (AI) has a significant contribution in areas of vaccine discovery. The rise of ML has greatly accelerated the often lengthy vaccine approval process. ML models for COVID-19 vaccine development focus on the prediction of potential epitopes by using a variety of techniques, such as artificial neural network, gradient boosting decision tree and deep neural network. In this regard, immuno-informatics tools are time-saving and cost-effective methods to hasten the design and establishment of a proficient multi-peptide candidate vaccine. The utilization of multi-epitope-based vaccines has been demonstrated to be a promising immunization approach against viruses due to the induction of long-term protective immunity. Methods: In the present study, a comprehensive computational and machine learning based approach was conducted to design a multi-epitope-based potential candidate vaccine composed of cytotoxic T lymphocyte (CTL) and helper T lymphocyte (HTL) epitopes of conserved regions of Spike and Nucleocapsid proteins. The potential viral peptides as the candidate vaccine were screened regarding convenient features like hydrophilicity, flexibility, antigenicity, and charging properties. In the next step, the vaccine efficacy needs to be improved by an immune adjuvant. For this purpose, the C-terminal domain of the heat shock protein gp96 (CT-GP96) was applied as a potent adjuvant for enhancing immunity. The five assembled constructs with different peptide orders were generated and fused with the assistance of suitable linkers. All five assembled candidate vaccine constructs were modeled and their 3D structures were assessed in terms of strong immune responses stimulation and their structural stability and immune processing for response induction. Finally, the best refined model was docked to toll-like receptor 4 (TLR4). Furthermore, Molecular Dynamics (MD) simulation of the vaccine receptor complex was done to assess the stability and related physical movements of the vaccine-TLR4 docking complex. The final candidate vaccine was computationally cloned in E. coli expression host to guarantee its high level production. Results: Following a comprehensive immune-informatics and machine learning based approach, the best conserved CTL and HTL immune stimulant epitopes were selected and assembled in different orders to build five different constructs. The final validated candidate vaccine construct was selected according to its efficacy, stability, and exposure ability, molecular docking analysis with TLR4. The molecular simulations by iMODS software also confirmed the stability of the binding interface. Additionally, the computational cloning of the final assembled candidate vaccine with pET28a plasmid showed the possibility of high level production of the vaccine construct post transformation in an E. coli host. Conclusion: The computational analysis indicated that this construct can be proposed as a potent prophylactic and therapeutic candidate multi-epitope vaccine against SARS‐COV‐2 once its effectiveness is verified by experimental and animal studies.