Аннотация
Emerging and re-emerging viruses are difficult to anticipate because epidemic potential emerges from interacting molecular, ecological and social processes. This narrative review critically synthesises evidence published from 2017 to July 2026 on entry, replication, assembly, innate immune evasion, genomic plasticity and cross-species adaptation across major virus families. Rather than treating sequence features as deterministic predictors, we distinguish mechanistic plausibility from phenotypic and epidemiological validation. We compare conserved vulnerabilities and evidence gaps across coronaviruses, orthomyxoviruses, filoviruses, arenaviruses, nairoviruses, hantaviruses, phenuiviruses, henipaviruses, flaviviruses, alphaviruses and orthopoxviruses. We then connect structural biology, metagenomic sequencing, CRISPR diagnostics, genomic epidemiology, artificial intelligence, vaccine platforms, antibodies and antivirals within an original sequence-to-decision framework. The synthesis shows that receptor compatibility is necessary but insufficient; polymerase activity, intracellular host dependence, immune antagonism and ecological exposure jointly shape risk. It also identifies persistent limitations: biased sampling, weak phenotype linkage, model-system uncertainty, unequal genomic capacity and limited countermeasure evidence for neglected pathogens. A systems-virology approach can improve prioritisation when predictions are explicitly graded, experimentally validated and embedded within equitable One Health surveillance.