This release provides the main script and resources for simulating genotypes from a coalescent model with samples from four populations, aimed at studying the effects of population structure on regression-based epistasis detection. The pipeline implements an analysis both with and without PCA-based correction for population structure. The results highlight how population structure can lead to inflated false positives in genome-wide association studies (GWAS) that model epistasis as pairwise interactions.