Background: Neisseria gonorrhoeae remains a major global public health challenge due to increasing antimicrobial resistance, limited treatment options, and the absence of an effective vaccine. The rapid emergence of drug-resistant gonococcal strains underscores the urgent need for accelerated vaccine development. In this study, we integrated reverse vaccinology with artificial intelligence–driven immuno-informatics to systematically identify conserved, immunogenic, and surface-accessible antigens for rational vaccine design against N. gonorrhoeae.
Methods: The complete proteome of N. gonorrhoeae FA1090 was screened using a multi-step reverse vaccinology pipeline. Subcellular localization, adhesin probability, signal peptides, transmembrane helices, essentiality and host homology were evaluated to filter for non-human, surface-exposed, and biologically relevant proteins. AI-based epitope prediction models, including deep-learning networks for B-cell and T-cell epitope identification, were employed to evaluate immunogenic potential. MHC class I and II binding affinity prediction was performed using machine learning covering prevalent human and murine alleles to ensure translational suitability for downstream preclinical validation.
Proteins conserved across global N. gonorrhoeae isolates were prioritized through pan-genomic conservation analysis. Structural modelling and molecular docking were employed to assess epitope accessibility and stability. Antigenicity, allergenicity, and toxicity filters ensured immunological safety. Systems immunology scoring integrated AI-derived features to rank the most promising vaccine candidates.
Results: This integrated approach identified a shortlist of high-value antigens enriched for outer-membrane proteins, transporters, and virulence-associated factors. Along with some novel candidates that are obtained from the hypothetical proteins of N. gonorrhoeae some other notable proteins such as BamA, LptD, MtrE, and NGO2054 consistently demonstrated high antigenicity, surface exposure, broad conservation and strong AI-predicted epitope profiles. Several predicted epitopes exhibited high binding affinity across multiple MHC class I and II alleles, supporting their potential utility in both human populations and mouse models. Structural analysis confirmed the top-ranked epitope clusters are surface-exposed and spatially accessible for antibody recognition.
Conclusion: Our findings underscore the power of integrating reverse vaccinology with AI-driven immuno-informatics to accelerate the rational discovery of gonococcal vaccine targets. The shortlisted antigens and epitopes represent promising candidates for peptide-based or multi-epitope vaccine strategies against resistant N. gonorrhoeae strains. Further experimental validation is underway to evaluate their immunogenicity and protective potential.