Fig. 3

Different branching pattern in the nucleotide and protein trees caused by a compensatory frameshift. a A compensatory frameshift was identified in the SEOV nucleocapsid protein. There were 46 reference and two compensatory frameshift form sequences. b A part of the multiple alignment of representative SEOV nucleocapsid CDS sequences is shown. The compensatory frameshift form sequences (in red) have a 1-nt deletion (open triangle) and a 1-nt insertion (filled triangle). The alignment of all 48 nucleotide sequences is presented in Additional file 3: Fig. S4. c A part of protein sequence alignment shows the 7-aa frameshifted sequences (in red). d, e Phylogenetic trees inferred from full-length nucleotide (d) and protein (e) sequences are presented. The ancestral branch of the compensatory frameshift form sequences (in red) is indicated by a red arrowhead. f, g Phylogenetic trees inferred from nucleotide (f) and protein (g) sequences without the frameshift segment revealed that the compensatory frameshift form clade (in red) was placed among reference form sequences in both trees. High resolution images of phylogenetic trees (d–g) are presented in Additional file 3: Fig. S5. h Protein distance (Dp) values between the reference sequence and all the other sequences were plotted against corresponding nucleotide distance (Dn) values. Dp values of the compensatory frameshift form sequences (red dots) were higher compared to those of reference form sequences (black dots) albeit their similar Dn values. i When the frameshifted segment was removed, Dp values of the compensatory frameshift form sequences were similar to those of the reference form sequences