Genome-wide amino acid substitution dynamics reveal conserved functional constraints and candidate variable sites in human group A rotavirus

Human group A rotavirus (RVA) remains the leading cause of severe acute gastroenteritis in children under 5 years of age worldwide. Although previous studies have examined evolutionary pressures acting on individual RVA genes, the physicochemical nature of amino acid substitutions influencing protein structure, function, and viral fitness has not been systematically investigated. Here, we performed a genome-wide analysis of amino acid substitution dynamics across all 11 RVA proteins to identify conserved functional regions and candidate variable sites. Coding sequences for each RVA gene were retrieved from the NCBI database, aligned using a codon-based strategy, and translated into protein sequences. Amino acid substitution matrices were generated for each protein, and substitutions were classified as conservative or radical based on physicochemical residue classes. Positional variability scores corresponding to the Shannon entropy thresholds were used to identify conserved residues and highly variable sites, which were mapped onto experimentally characterized structural and functional domains. Amino acid substitutions across the RVA proteome exhibited highly asymmetric and non-random patterns, with conservative substitutions predominating over radical substitutions. Hydrophobic-to-hydrophobic replacements were the most frequent, whereas substitutions involving acidic and basic residues were comparatively rare. The internal structural proteins VP1, VP2, and VP6, together with NSP2 and NSP5, displayed highly constrained substitution profiles, with VP6 being the most conserved protein. In contrast, the outer capsid proteins VP4 and VP7 and the non-structural proteins NSP1 and NSP4 exhibited substantially greater substitution diversity. Critical functional motifs, including the VP1 RNA-dependent RNA polymerase GDD motif, VP1–VP3 interaction domain, VP2 RNA-binding residues, VP3 2'-O-methyltransferase domain, and the NSP5 phosphorylation region, remained conserved, whereas the core structural domain of VP7, the VP8*/VP5* domains of VP4, and the variable core domain of NSP1 contained the highest density of variable residues. These findings demonstrate that RVA evolution is governed by a balance between sequence diversification and functional constraint, in which conservative substitutions preserve protein integrity while radical substitutions are largely confined to surface-exposed regions, consistent with the hypothesis of immune-mediated selection. Conserved domains identified in this study may represent promising candidates for broad-spectrum vaccine and diagnostic development, whereas hypervariable regions provide valuable markers for molecular surveillance of RVA evolution.