Synonymous mutations in viral evolution and vaccine design

Synonymous mutations, long regarded as evolutionarily neutral, are increasingly recognized as important determinants of viral evolution and adaptation. In this review, we propose that viral codon usage reflects a multi-objective optimization process shaped by competing evolutionary constraints, positioning synonymous variation as a key layer of translational control with direct implications for viral evolution and rational vaccine design. We first summarize the quantitative metrics used to characterize viral codon usage bias and then examine the evolutionary forces that shape viral codon usage profiles. In particular, we discuss growing evidence that viral codon usage frequently diverges from host preferences, reflecting mutational bias, trade-offs among translational efficiency, immune pressure, RNA structure, co-translational protein folding, and transmission between hosts, rather than simple optimization. We then focus on strategies that deliberately manipulate codon usage for vaccine development. Finally, we discuss the current challenges and how emerging artificial intelligence-based approaches may enable more predictive and integrative codon engineering strategies.