2.3 Phylogenetic and evolutionary analyses
Eleven full-length genome sequences derived from this study were analyzed and compared with the dataset of 188 full-length genome sequences of CPV-2, originally isolated from domestic dogs during 1979 to 2017, available in the GenBank database (Supplementary Table S4). The substitution rate of the full-length genome sequences and each individual structural and nonstructural gene coding region was estimated as the number of nucleotide substitutions per site per year (nt/s/y), implemented in the Bayesian Evolutionary Analysis Sampling Trees (BEAST) version 1.8.4 software (http://tree.bio.ed.ac.uk/) . The rates were calculated based on the two main phylogenetic clades of the CPV-2 strains discovered in Asia (Asian clade) and the CPV-2 strains reported in Europe and America (Western clade). The best-fit model of substitution (HKI + I + G4) was determined using the model section function in the MEGA 7.0 software (https://www.megasoftware.net/). The coalescent Bayesian skyline was applied under exponential relaxed-clock models (Drummond & Rambaut, 2007; Li et al., 2017). The Markov Chain Monte Carlo (MCMC) algorithm was run for 200-million generations and logged every 10,000 states. The convergence of all parameters was checked by observing that the effective sample size (ESS) > 200 with TRACER version 1.6.1. The maximum clade credibility (MCC) tree was obtained from the posterior distribution of trees using TreeAnnotator version 1.8 after discarding 10% of steps as burn-in. The phylogenetic tree with timeline estimated divergences, posterior probability (PP) and 95% highest probability density (HPD) values were generated using FigTree version 1.4.2.