Instructor: Josef C Uyeda (josef.uyeda@gmail.com)
This repository provides material for a bayou tutorial at the 2017 Evolutionary Quantitative Genetics workshop at Friday Harbor Laboratories. The total exercise will likely take substantially longer than we have, but we'll get as far as we can, with special emphasis on using priors to constrain OU models to be more like our microevolutionary data. Before attending the workshop, please make sure you have the needed software and packages installed. Instructions can be found on the course Wiki.
The R package bayou allows to use Bayesian modeling of adaptive trait evolution on phylogenies using Ornstein-Uhlenbeck (OU) models. These models can be used to a test a variety of evolutionary questions regarding the tempo and mode of trait evolution.
-
We will learn the theory behind the use of Ornstein-Uhlenbeck models in phylogenetic comparative approaches.
-
Gain an intuitive sense for how OU models work and how they relate to other models
-
Understand how the parameters of the model relate to biological processes, and how they are estimated
-
Obtain a basic understanding of Bayesian reversible-jump MCMC
-
-
We will discuss strategies for using bayou to answer biological questions, and how the approaches implemented in bayou relate to other implementations of OU models.
-
Attendees will obtain hands-on experience using bayou to:
-
Identify the location and magnitude of adaptive shifts using reversible-jump MCMC
-
Diagnose, adjust and tune an MCMC
-
Perform Bayesian model selection in a hypothesis-testing framework
-
Implement customized regression models in an OU modeling framework
-