Authors: Baylee Christensen, Reagan Mckee, Dante Celani, Candice Johnson, Randy Klabacka
Mentors: Randy Klabacka
Insitution: Utah Tech University
When studying the genetics of biological systems, we often assign individuals to categories (e.g., “ecotypes”) and then assess genetic differences using computational biology tools. If populations within categories are used as units in statistical models, this can present a potential statistical pitfall called pseudo-replication (which happens when multiple measurements of the same population are compared). To circumvent this flaw, we developed a software package called CatPop. CatPop considers all possible population assignments (using combinatorics) and performs a permutation test to determine whether a locus has significantly greater divergence between groups compared to within the same group. Our test of CatPop on simulated data shows that it can accurately identify divergent loci between categories, and we also examine its utility with previously-published datasets that examine divergence between ecotypes.