Choose timezone
Your profile timezone:
Description
Inferring fitness from genomic sequences is a central challenge in evolutionary populations and infectious disease surveillance. This talk presents related studies on reconstructing fitness landscapes from whole-genome, time-stratified population data. Using SARS-CoV-2 as a case study, we compare transient Quasi-Linkage Equilibrium (tQLE), which incorporates epistasis under a near-linkage-equilibrium assumption, with Maximum Path Likelihood (MPL), which assumes additive fitness but accommodates arbitrary allele correlations. We further examine, through theory and simulations, the parameter regimes in which genotype fitness order can be reliably inferred under selection, mutation, and recombination. We extend the QLE theory to interacting populations connected by migration and show that low migration rates preserve the QLE phase, enabling accurate inference of additive and epistatic fitness parameters. The studies clarify both the potential and the limitations of fitness inference from large-scale genomic time-series data.