Zaragoza, España
In this study we introduce the spectral radius of the molecular coancestry matrix, computed for an enclosed set of contiguous markers, as a selection signature test. We emphasize that this test is useful when the sampled individuals are related (a typical situation in animal breeding) as compared to other tests that rest on the assumption that sampled sequences are randomly drawn. This is because the genomic relationship matrix intrinsically accounts for relatedness by construction. We explore the statistical power of the spectral radius test versus other standard neutrality tests through a stochastic simulation study under several population scenarios. We found that the test based in the spectral radius outperformed the Tajima’s test in all the scenarios assayed. In turn, the power was only greater than the one obtained with the rEHH statistic for certain specific situations, noticeably when the frequency of the causal mutation at the start of the selection is low and subsequent selection intensity is strong. We conclude that the method is comparatively useful to detect intense selective sweeps.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados