Rethinking classical swine fever virus phylogenetics: NS3 Outperforms traditional gene markers

Classical swine fever virus (CSFV) remains a major threat to global swine production due to its high transmissibility, genetic diversity, and continued outbreaks despite vaccination and control programs. Although whole-genome sequencing provides the most reliable framework for phylogenetic inference and molecular epidemiology, its routine use is often constrained by cost, infrastructure, and processing time. Identifying sub-genomic regions that reliably represent whole-genome evolutionary relationships is therefore essential for effective surveillance and outbreak investigations. The aim of this study was to identify whole-gene regions of the CSFV genome that most accurately recapitulate whole-genome phylogenetic patterns and can serve as robust alternatives to complete genome sequencing. A total of 67 complete CSFV genomes representing all major genotypes and subgenotypes were analyzed. Whole-genome and gene-wise correlation analyses of pairwise genetic distances were performed, along with comparative phylogenetic reconstruction using a maximum-likelihood framework. Among all genomic regions examined, the non-structural genes NS3, NS5A, and NS5B showed substitution model concordance with the whole-genome and exhibited the strongest correlations with whole-genome genetic distances. Notably, NS3 demonstrated the highest concordance and generated phylogenetic trees that most closely mirrored whole-genome topology. In contrast, the commonly used E2 gene displayed reduced concordance, indicating limitations as a standalone phylogenetic marker. Overall, this study identifies NS3 as the most reliable whole-gene surrogate for CSFV whole-genome phylogenetic inference, providing a practical framework for accurate molecular epidemiology and enhanced CSFV surveillance.