Genetic vaccine performance is determined not only by the delivery platforms and antigen choice but also by the coding sequence that influences antigen expression, innate immune sensing, transcript stability and sequence-associated risks. However, the design principles governing these sequence-level effects remain insufficiently defined. Thus, we consider vaccine coding-sequence design as a multi-objective engineering problem that requires simultaneous consideration of multiple, potentially competing biological properties. We present a multi-parameter computational framework for analysing SARS-CoV-2 vaccine coding sequences, including BNT162b2 (Pfizer-BioNTech), mRNA-1273 (Moderna), AZD1222 (AstraZeneca) and several academic vaccine candidates. The framework evaluates codon usage, codon-pair context, CpG content, regulatory motifs, GC content and localized sequence-structural features within a unified quantitative design space. Our analysis shows that existing vaccines use diverse design strategies and that no single metric adequately captures their design complexity. By enabling systematic comparison across multiple parameters, the framework supports the rational generation and prioritization of candidate sequences with defined biological trade-offs. This work bridges computational sequence analysis, vaccine biology and engineering design, providing a foundation for context-aware optimization of genetic vaccine sequences across delivery platforms, target tissues and emerging pathogens.
