Since some common approaches to the study of molecular adaptation may not be optimal for answering questions regarding within-host virus evolution, we have developed an alternative approach that estimates an absolute rate of molecular adaptation from serially-sampled viral populations. Here, we extend this framework to include sampling error when estimating the rate of adaptation, which is an important addition when analyzing historical data sets obtained in the pre-HAART era, for which the number of sequences per time point is often limited. We applied this extended method to a cohort of 24 pediatric HIV-1 patients and discovered that viral adaptation is strongly associated with the rate of disease progression, which is in contrast to previous analyses of these data that did not find a significant association. Strikingly, this results in a negative relationship between the rate of viral adaptation and viral population size, which is unexpected under standard micro-evolutionary models since larger populations are predicted to fix more mutations per unit time than smaller populations. Our findings indicate that the negative correlation is unlikely to be driven by relaxation of selective constraint, but instead by significant variation in host immune responses. Consequently, this supports a previously proposed non-linear model of viral adaptation in which host immunity imposes counteracting effects on population size and selection.
Raghwani J, Bhatt S, Pybus OG (2016) Faster Adaptation in Smaller Populations: Counterintuitive Evolution of HIV during Childhood Infection. PLoS Comput Biol 12(1): e1004694. http://dx.doi.org/10.1371/journal.pcbi.1004694 ;