Figure 3

HIV-1 demonstrates higher G-to-A hypermutant frequencies than HIV-2. a The frequencies of each type of transition hypermutant were compared between HIV-1, HIV-2, and the plasmid controls. For this analysis, hypermutants were defined as read pairs containing two or more mutations of the indicated type within an individual read pair (approximately 120 bp in length). The frequency of hypermutation was then calculated by dividing the number of hypermutant read pairs by all read pairs. b The frequency of G-to-A hypermutation was compared across all five amplicons examined by Illumina DNA sequencing. c The degree of G-to-A hypermutation was analyzed by determining the numbers of G-to-A mutations within hypermutant read pairs. d The dinucleotide context of G-to-A mutations from G-to-A hypermutants was determined and expressed as a percentage of the total. Data in panels a, b, and d represent the mean of three experimental replicates, with error bars representing standard deviation, while data in panel c represent the total (i.e. compiled) data. Asterisks denote statistically significant differences between HIV-1 and 2 (*p < 0.05, ***p < 0.001, N.S. not significant). The actual numbers of G-to-A hypermutant read pairs and reference read pairs are listed in Additional file 6: Dataset S2.