Figure Legends

Figure 1. Reassortant detection analysis. Three reassortants were found by RDP4 (Martin et al., 2015) among the 17 TiLVs analysed, namely TH-2016-CN, IL-2011 and TH-2018-K. The inferred breakpoints of conflicting evolutionary signals appeared to match well with the segment boundaries (vertical dotted lines). The segment numbers are shown below the diagram. The analysis suggested that segments 1–4 shared the same evolutionary history. Segment 5 and 6 had their own unique histories. Segments 7–10 shared the same history.
Figure 2. Evolutionary dating analyses . Time-calibrated phylogenies for four genomic regions of the 17 TiLVs were inferred using BEAST v1.10.4 (Rambaut et al., 2018). See Methods and Materials for more details. Node bars represent 95% HPD intervals of date estimates. The numbers on nodes are posterior clade probabilities, shown for those with values > 0.80. Tips are coloured according to their countries of origin (black, Thailand; blue, Bangladesh; red, USA; yellow, Israel; green, Peru; and orange, Ecuador).
Figure 3. Demographic history of TiLV. The past population dynamics of the 17 TiLVs was inferred for various genomic regions by using the GMRF Bayesian Skyride coalescent model, implemented in BEAST v1.10.4 (Rambaut et al., 2018). Solid line is the posterior median of the effective population size through time. Dashed lines represent the 95% HPD interval. The y-axis shows the log-scaled effective population size Neτ where Neis the effective population size and τ is the viral generation time in years.