Work in our lab is united by a central focus on evolutionary biology and conservation biology. Some of our current projects focus tightly on computational biology and statistical methods in phylogenomics, some of them focus on empirical work (mostly in amphibians and reptiles), and some of them focus on translating science into conservation. Our list of active projects is always changing, but a few recent examples are listed below.
Phylogenomics: Model performance and enhancing accuracy of analysis
In collaboration with Jeremy Brown’s lab (LSU), we are working on an NSF funded project developing absolute measures of model performance. The approach we’re taking is rooted in posterior predictive simulation, which seeks to answer the question ‘Can the model we’re using reasonably predict the data that we’re using it to analyze?’ Put another way, does our model capture the important sources of variation in the data?
We can contrast this with measures of relative model performance, such as model selection using AIC. In this case, the aim is to choose a best fitting model from a set of candidate models, rather than ask if the model fits the data well in an absolute sense. This distinction becomes important because it is possible that all models in the set of candidates are a poor fit to the data, so the best fitting one may nonetheless provide biased answers. This is especially likely to happen for large scale ‘phylogenomic’ datasets that have become standard in the field.
This project is part of a more general interest in good practices for phylogenomic analyses, assessing accuracy of phylogenetic and species delimitation methods, and enhancing reproducibility of analyses.
Phylogenetics and Systematics of Turtles
Much of our work focuses on resolving phylogeny and species boundaries of turtles. Along with our collaborators, we have developed phylogenies for a large fraction of the turtle species in the world. These projects include the first attempts at assembling a comprehensive turtle tree of life (Iverson et al. 2007, Thomson and Shaffer 2010), as well as deeply sampled phylogenies focusing on especially problematic clades (e.g., the Map Turtles of the genus Graptemys) and, most recently, a large analysis of 577 turtles from 288 species. This is the closest to a comprehensive phylogeny that we currently have for turtles and is the result of roughly 10 years of hard work through the NSF-funded ‘Turtles of the World’ project. Another major effort in this vein was sequencing the first turtle genome which was completed in 2014. Bob served on the project’s steering committee and led the phylogenetic portion of the project.
Whiptail lizards: Species boundaries and complex evolutionary histories