Molecular phylogenetic analysis of bovine viral diarrhoea virus: a Bayesian approach

Genetic typing of bovine viral diarrhoea virus (BVDV) is important for precise classification of viruses. Traditionally, inferring BVDV phylogeny has been performed by distance-based method, i.e. neighbor-joining for single genes. In this study, a Bayesian approach was exploited to analyze five genetic regions of BVDV genome (5' UTR, N(pro), E2a, E2b, and NS3) for 68 taxa retrieved from GenBank. The results showed that all taxa in the consensus tree of E2a have been assigned correctly to corresponding groups, i.e. type-2 BVDV, and BVDV-1a, -1b, -1c, -1e, and -1g, supported by a high posterior probability. In contrast, subgroup 1a formed polytomies in the consensus trees of 5' UTR and NS3. Polytomies also appeared among the subgroup 1b in the consensus trees of N(pro) and E2b. Analysis of a combined dataset produced an unambiguous, well-supported phylogenetic hypothesis. The topologies found for each genetic region separately and combined were different, but the differences were mainly weakly supported by the data. Combining the data allowed the identification of well-supported clades of strains that correspond to some of the previously defined subgroups. Only a combined approach will allow the confident placement of new strains in the current classification of viruses into genotype and subgenotype.