|Home | About | Journals | Submit | Contact Us | Français|
The complex evolutionary process of human immunodeficiency virus type 1 (HIV-1) is marked by a high level of genetic variation. It has been shown that the HIV-1 genome is characterized by variable and more constant regions, unequal nucleotide frequencies, and preference for G-to-A substitutions. However, this knowledge has largely been neglected in phylogenetic analyses of HIV-1 nucleotide sequences, even though these analyses are applied to a number of important biological questions. The purpose of this study was to identify a realistic model of HIV-1 evolution and to statistically test if the application of such a model significantly improves the accuracy of phylogenetic analyses. A unique and recently reported HIV-1 transmission cluster consisting of nine infected individuals, for whom the direction and time for each transmission were exactly known, formed the basis for the analyses which were performed under a general model of nucleotide substitution using population sequences from the env V3 and p17gag regions of the HIV-1 genome. Examination of seven different substitution models by maximum-likelihood methods revealed that the fit of the general reversible (REV) model was significantly better than that of simpler models, indicating that it is important to account for the asymmetric substitution pattern of HIV-1 and that the nucleotide substitution rate varied significantly across sites. The shape parameter alpha, which describes the variation across sites by a gamma distribution, was estimated to be 0.38 and 0.25 for env V3 and p17gag, respectively. In env V3, the estimated average transition/transversion rate ratio was 1.42. Thus, the REV model with variable rates across sites (described by a gamma distribution) provides the best description of HIV-1 evolution, whereas simple models are unrealistic and inaccurate. It is likely that the accuracy of phylogenetic studies of HIV-1 and many other viruses would improve substantially by the use of more realistic nucleotide substitution models. This is especially true when attempts are made to estimate the age of distant viral ancestors from contemporary viral sequences.