Predicting viral evolution with memory-driven models

Diseases like influenza or SARS-COV-2 evolve so quickly that forecasting future viral variants remains a major challenge in vaccine development. Traditional approaches estimate how these variants spread through the population using models that treat each generation as though it is independent of its past. Yet, emerging research suggests that the history of how a virus changes can linger, shaping future trajectories in subtle yet powerful ways. By factoring in these long-lived “memory” effects, scientists can hope for a better understanding of why certain viral strains suddenly dominate, ultimately improving predictions for the next wave of infections.

In this project, you will investigate how best to capture these memory signals in genetic frequency data using mathematical random walk models. The work will touch on mathematical modelling, programming simulations and analyzing real or simulated datasets, offering a glimpse of how quantitative tools might transform strategies for vaccine design and disease control. No biological experience is required. Experience in a programming language like Python, R or MATLAB experience is preferred. Experience in mathematical modelling is preferred.