Artificial intelligence, structural, and immunoinformatics-driven multi-epitope vaccine design targeting...

West Nile Virus (WNV), a globally distributed mosquito-borne flavivirus, poses a significant public health threat due to its potential to cause severe neurological complications and the absence of licensed vaccines or specific antiviral treatments. In this study, we applied an integrated artificial intelligence, structural, and immunoinformatics-driven approach to design a multi-epitope subunit vaccine (MESV) targeting the non-structural proteins NS1 and NS4B, which play key roles in viral replication and immune evasion. B-cell, cytotoxic T lymphocyte (CTL), and helper T lymphocyte (HTL) epitopes were computationally predicted and screened for high antigenicity, non-allergenicity, and non-toxicity. Selected epitopes were assembled into a 307-amino-acid construct incorporating the 50S ribosomal protein L7/L12 and RS09 as adjuvants, along with a PADRE sequence to enhance T-helper responses. Structural modeling, physicochemical analysis, and mRNA structure prediction confirmed stability, solubility, and immunogenic potential. Molecular docking and 100-ns molecular dynamics simulations demonstrated strong, sustained interactions with TLR3 and TLR4, with principal component analysis supporting conformational stability. Immune simulations predicted robust humoral and cellular responses, including elevated IgM, IgG1, IgG2, cytokine production, and memory formation. Population coverage analysis revealed broad HLA representation, particularly in Europe (99.55%). Codon optimization and in silico cloning confirmed suitability for E. coli expression. Collectively, these AI-assisted computational insights highlight the MESV as a promising WNV vaccine candidate and provide a rational framework for future experimental evaluation.

The online version contains supplementary material available at 10.1007/s40203-025-00459-6.