Predicting HIV-1 coreceptor usage with sequence analysis.

Bioinformatics approaches are increasingly being used to identify and understand the genetic variation underlying changes in HIV-1 biological phenotype. The variable regions of the viral envelope are the major determinant of virus coreceptor usage and cell tropism. Specifically, amino acids 11 and 25 in the 3rd variable (V3) loop have been found to strongly influence viral syncytium inducing capacity and coreceptor usage. Many additional V3 loop changes, however, as well as changes elsewhere in Env, are thought to contribute to phenotype. In this review we describe several recently developed methods to analyze this variability and their use to predict biological phenotype based on sequence information. These approaches have identified changes in the V3 loop, in addition to the known changes at positions 11 and 25, that affect phenotype and significantly enhance our ability to predict phenotype from genotype. Besides improving phenotype prediction, methods that score V3 sequences on a continuous scale can also assist in the interpretation of evolutionary information about shifts in phenotype, and the relationship between that evolution and pathogenesis. Several examples and potential practical applications of this scoring are discussed. We conclude that advances in computational approaches have enhanced both our ability to predict and to understand HIV-1 biological phenotype evolution. Further development of these methods, by extending analysis to regions outside the V3 loop and to clades beyond subtype B, will extend our understanding of HIV-1 pathogenesis and inform treatment strategies.