DEEPR: A Dirichlet-multinomial randomization test for estimating relative coevolutionary event differences between groups of symbiotic species
Year: 2014
Type: dissertation
Abstract: This thesis presents a novel statistical method for investigating differences in the coevolutionary history of symbiotic organisms, with a focus on cospeciation and host switch events. Past research has focused on testing for significant congruence and incongruence within single cophylogenies, the former being evidence of cospeciation, with the latter suggesting host switching. However, existing methods lack the capability of describing differences between cophylogenies, or summarizing groups/taxa in which different degrees of congruence may be found. Here, I present DEEPR (Dirichlet-multinomial Evolutionary Event Profile Randomization test), which fits Dirichlet-multinomial models to groups of cophylogenies and compares them through a randomization test. DEEPR is validated with artificial data generated under known parameters, and applied to a real biological example comparing virus and bacterial symbiont phylogenies. This novel test opens new research avenues in comparative coevolutionary biology.
Author: Mark Merilo
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W2606026853