Beyond one drug–one target: AI-guided discovery of multi-target antiviral therapeutics

Abstract

Viral infections remain a major health problem in the world because they have high mutation rates, are highly adaptable and they develop resistance to traditional antiviral drugs. The conventional one drug, one target approach, though useful in some situations is getting more and more constrained by the development of resistant viral strains and partial blocking of intricate viral life cycles. Multi-target antiviral agents (MTAs) have been identified in this respect as a promising therapeutic option with the ability to simultaneously regulate a variety of viral and host-related pathways. This review thoroughly examines the theoretical framework, rationalization, and therapeutic benefits of the MTAs and their potential to improve antiviral effects, expand spectrum, and inhibit the development of resistance. Moreover, the current review article highlights how artificial intelligence (AI) and machine learning (ML) can be used to speed up multi-target drug discovery in a transformative way. The high-performance computational methods, such as structure-based and ligand-based design, molecular docking, pharmacophore modeling, network pharmacology and generative AI, allow candidates to be found and optimized rapidly to multi-targets. Medicinal chemistry approaches, including hybrid pharmacophore design, molecular conjugation, and host-targeted, and the role of natural product-derived scaffolds in the development of multi-target antiviral agents are also discussed in the review. The major case studies of HIV, hepatitis C virus, SARS-CoV-2, and influenza are discussed as an example of translational relevance. Lastly, the critical issues under the pharmacokinetics, toxicity, and clinical development are discussed, and the future outlook is on AI-based precision antiviral therapeutics. Taken together, this review article highlights the paradigm shift to AI-driven multi-target discoveries of next-generation antiviral drugs.