Fig. 4

a TuMV genome organisation. b Nucleotide distance vs amino acid distance of the TuMV. Slopes values for each regression are indicated. All regression p < 0.001. c Nucleotide vs amino acid pairwise distances for TuMV genes. For each gene, the Huber estimator (±median absolute deviation) of the distances normalised to their respective concatenated is displayed. d Hierarchical cluster dendogram using the eleven chosen variables (see Additional file 5: Table S3a) for TuMV. The clustering was performed using the euclidean distances and the Ward method. Probability values were calculated using bootstrap resampling techniques, the approximately unbiased (AU) support p-value (red) and the bootstrap probability (BP) value (red). e Biplot of the principal component analysis (PCA) using the eleven chosen variables (see Additional file 5: Table S3a) for TuMV. The first principal component is represented in the x-axis, and the second principal component is represented in the y-axis. Percentage values in the axes indicate the percentage of variation explained by either components. Original variables are given in blue, and those showing co-variation above 0.8 are encircled by discontinuous lines