EX 527 Analytical models of HIV-1 evolution with recombination allow description of viral evolution not only at the extremes of very small and very large Ne, where drift and selection, respectively, dominate, but also at intermediate values of Ne where both selection and drift remain important simultaneously. The models, however, are restricted to a small number of loci and/or to simple fitness landscapes. Experimental data on viral diversification, in contrast, is available over genomic regions that are up to several hundred nucleotides long. Besides, the best available description of the HIV-1 fitness landscape points to significant deviations from a simple multiplicative fitness profile. To overcome these limitations of analytical models, we have recently developed bit-string simulations of the within-host genomic diversification of HIV-1. Our simulations consider large genome lengths and incorporate mutation, infection of cells by multiple virions, recombination, fitness selection, and epistatic interactions between multiple loci, thereby presenting a detailed description of the evolution of viral diversity and divergence in infected individuals. In particular, our simulations elucidate the role of recombination in HIV-1 diversification as a function of Ne and with the experimentally determined fitness landscape. Here, we apply the simulations to describe patient data and obtain estimates of Ne. We examine the dependence of our estimates of Ne on the frequency of multiple infections of cells and on the nature of the fitness landscape, which remain to be established in vivo. Finally, we revisit the large estimate of Ne obtained by Rouzine and Coffin by incorporating multiple infections of cells and recombination in their two-locus/two-allele model. We perform simulations to predict the evolution of viral diversity, dG, and divergence, dS, in an HIV-1 infected individual. Diversity is a measure of the genomic variation in the viral population at any given time, whereas divergence is a measure of the deviation of the viral genomes from the founder strain. Simulations begin with the synchronous infection of a fixed population, C, of cells by identical homozygous virions. Following infection, viral RNA are reverse transcribed to proviral DNA, during which FTY720 company process mutation and recombination introduce genomic variation. Proviral DNA are then transcribed into viral RNA, which are randomly assorted into pairs and released as new virions. New virions are chosen according to their fitness to infect the next generation of uninfected cells, and the cycle is repeated. We employ parameter values representative of HIV-1 infection in vivo.