Species evolution


Recently algorithms for estimating species trees, rather than gene trees, have initiated a paradigm shift in evolutionary biology that is clarifying many issues in the study of phylogenetics, but also raising new conceptual challenges. Our centre has conducted research in this area for a number of years, mostly based on implementing and elaborating the multispecies coalescent model for embedding gene trees inside a shared species tree. 

Genomes, phenotypes and fossils: integrative models of species evolution


Bayesian Integrative Models

How and when species came to be is the fundamental question in macroevolution. Attempts to answer it use a variety of data sources including genome sequences, morphology and fossil discoveries. Yet current methods are unable to exploit all this data, with different data sources often producing conflicting results. This research project aims to create a unifying probabilistic framework that combines genomic, fossil and phenotypic data to give us the best possible understanding of evolutionary history. The research will involve (i) developing new mathematical models, (ii) creating open-source software tools to disseminate new methods widely, and (iii) using these new methods to address outstanding questions in human, animal and pathogen evolution.

The initial focus of the work will be to extend the StarBEAST2 package to allow for sampled ancestral species and their phenotypes in the species tree as well as ancient DNA samples in the embedded gene trees. This is a major software engineering task. There will also be work on developing new trait evolution models that can account for trait variation both within and between species. Current models will be incorporated into the BEAST 2 software and major studies on real and simulated data will be run to assess their strengths and weaknesses. We will also implement new inference methods for correlated continuous trait evolution. Finally a third focus of the work will be on incorporating rich fossil data into the phylogenetic framework within BEAST 2. New models will incorporate variability of sampling over time and space, trait-dependent sampling, and will be able to use multiple fossils from the same morphospecies while accounting for uncertainty in the geologically-derived age of fossils. Both simulated and curated data sets will be used to test and prove the newly developed methods.

This is a three-year research project supported by the Marsden Fund starting 1 March 2017. This work is being undertaken by an international team lead by Professor Alexei Drummond and Dr David Welch (University of Auckland) with Associate Professor Tanja Stadler (ETH Zurich), Dr Nick Matzke (ANU) and Dr Tim Vaughan (University of Auckland). The project is also engaging expert collaborators with knowledge of specific paleontological and molecular data sets including Dr Mana Dembo and Dr Mark Collard (hominins; Simon Fraser University) and Dr Graham Slater (canids; University of Chicago).

 

For more information on this research project contact Professor Alexei Drummond (alexei@cs.auckland.ac.nz) or Dr David Welch (david.welch@auckland.ac.nz).

 

Two new PhD positions will be supported by this project. More information available here.