Inferring fitness seascapes from evolutionary histories.

Evolutionary dynamics are often conceptualized as walks on an adaptive landscape, with populations climbing toward optimal fitness peaks. However, environmental changes can transform static landscapes into dynamic "fitness seascapes" where natural selection fluctuates in time. Here, we used a path integral approach derived from statistical physics to reveal time-varying selection pressures from genetic sequence data. We found that constraints on the fitness seascape are determined by the screened Poisson equation, which also describes the screening of electric fields in media with mobile charge carriers. In our model, changes in mutation frequencies act as "charges" that reveal the underlying fitness seascape, which is analogous to an electrostatic potential. After validating our method in simulations, we applied it to study how HIV-1 evolves to escape immune control by T cells within individual hosts. Our analysis showed that the fitness benefit of immune escape declines as T cell responses approach their peak intensity, suggesting that functional exhaustion may impair the effectiveness of the immune response against HIV-1. Overall, our approach provides a general framework for capturing the complex dynamics of natural selection in rapidly evolving populations.