In Silico Saturation-Mutagenesis-Based Genomic Mutation Risk Assessment for Enterovirus B

Enterovirus B (EVB) is the most prevalent species of human enteroviruses, responsible for a wide range of diseases, including hand, foot, and mouth disease, viral meningitis, myocarditis, and neonatal sepsis, imposing a significant disease burden primarily on children. Coxsackievirus B (CVB1-6) and various echovirus (E) serotypes are the major serotypes of EVB. Since no antiviral drug or vaccine is available, it is important to strengthen monitoring, risk assessment, and early warning of genomic variations for EVB. CVB1, CVB3, E6, and E30 were selected as representative EVB serotypes for this study due to the availability of three-dimensional structures and their global prevalence. To evaluate the mutation effects of structural proteins on structural stability and receptor-binding affinity, computational saturation mutagenesis of EVB serotypes was performed using FoldX. Furthermore, based on data from deep mutational scanning for CVB3, a risk prediction model for EVB fitness was constructed by machine learning algorithms and applied to other EVB serotypes. Finally, we integrated three phenotypes-structural stability, receptor-binding affinity and fitness-to evaluate genomic variation risk of EVB and tracked the prevalence of high-risk mutants in natural viral sequences through molecular evolution analysis and mutation profiles. We identified the N-terminus and C-terminus of VP1 and the EF loop of VP2 as the EVB regions of highest genomic variation risk, and high-risk mutations had played significant roles in viral evolutionary history. These findings provide a framework for multi-phenotypic and multi-data approaches to viral risk assessment and offer insights to support the development of antiviral drugs and vaccines.