Statistical inference of selection coefficient from temporal allele frequencies
Sprache des Vortragstitels:
Sprache des Tagungstitel:
The focus is on estimating selection coefficients for a Wright-Fisher Model. Given the inherently intractable nature of the model's likelihood function, the research delves into various approximation techniques to improve the accuracy and efficiency of selection coefficient estimations. Adhering to the Bayesian paradigm, posterior distributions and evaluate uncertainty are constructed in the results. Furthermore, innovative methods are incorporated to summarize high-dimensional allele frequency data during the likelihood approximation process. The methods aim to provide valuable insights into evolutionary mechanisms and aid in developing more accurate models for estimating the parameters of evolutionary forces over time. Compared to existing methods, the novel approach exhibits promising performance on both simulated and real-world data sets.
Sprache der Kurzfassung:
Hauptvortrag / Eingeladener Vortrag auf einer Tagung