Collaborative Research: ABI Innovation: A Bayesian Evolutionary Tree Analysis Database

University Of California-Davis, Davis CA

Investigators

Abstract

This collaborative grant seeks to develop a prototype Bayesian Evolutionary Tree Analysis Database (BETAbase), an interactive repository for the output of Bayesian phylogenetic studies. Phylogenies-estimates of the genealogical relationships among species-provide an explicit historical perspective that critically informs research in a vast and growing number of scientic disciplines. Increasingly, phylogenies are inferred in a Bayesian statistical framework, as it permits estimation under more complex and realistic statistical models from larger and increasingly complex datasets, and allows uncertainty in the resulting estimates to be clearly identified and accommodated. Implementation of BETAbase will substantially advance the field of Bayesian phylogenetic inference and, by extension, the myriad fields of biological sciences (including conservation, medicine, and agriculture) that rely on phylogenetic information. It will promote more rigorous and consistent evaluation of published phylogenetic studies thereby improving the quality of decisions based upon those studies. BETAbase will also provide unique training opportunities by developing a web-based educational environment where researchers can gain experience and develop the skills to identify and address problems encountered in the course of Bayesian phylogenetic analysis. The output of Bayesian phylogenetic studies typically includes both log files with sampled trees and the other parameters of the phylogenetic model. The output accumulated over the last decade constitutes an untapped source of information with the potential to transform the field of Bayesian phylogenetics and associated areas of biological research. The BETAbase prototype attempts to harness this potential to: (1) promote more rigorous evaluation of empirical phylogenetic studies; (2) focus theoretical efforts on areas where existing phylogenetic methods are deficient; and (3) implement software tools that enable solicitation of empirically based informative priors that appropriately reflect our accumulated knowledge regarding model parameters, providing an alternative to the vague `uninformative' priors that are currently used as the default settings in existing Bayesian phylogenetic inference software. The project will involve the participation and training of two postdoctoral scholars as well as graduate and undergraduate students. Finally, BETAbase will safeguard the substantial computational investment that the systematic community has made toward understanding the Tree of Life. The prototype BETAbase archive and associated tools will be distributed and discussed at http://beta-base.org.