Diego Salazar - Evolutionary simulations to evaluate trait detection in GWAS: the effect of sample size and genetic architecture

Home institution and supervisors
Jorge Luis Ramirez Malaver. UNMSM, Peru.

Host institution and supervisors
José Cerca, Mark Ravinet, and Erik Sandertun Røed. UiO, Norway. 2024-2026.

Project description
This MSc project focuses on simulating population-level data under common scenarios of population split and divergence to test how can we improve the power of GWAS-based approaches. Using SLiM, an evolutionary simulation framework, two main scenarios will be coded: a single population where a trait arises and is maintained; and a two-population model where the trait evolves in one population and spreads to the second via gene flow. The goal of this research is to understand how a population's evolutionary histories and the characteristics of the trait influence the detection of genes of interest and to provide recommendations for conducting GWA studies. To this end, the genomic data will be generated while varying population sizes, gene flow ratios, selection coefficients, and the genetic architecture of the trait. This will ultimately enable assessing the effectiveness of common GWAS methods in detecting these traits with different sample sizes. Although each scenario is simple, the number of combinations grows rapidly when considering parameters such as selection coefficients, trait complexity, and GWAS methods. Lastly, the inferences from these simulations will be validated against empirical datasets such as Heliconius, three-spined sticklebacks, or Darwin’s finches.