Improving Bayesian methods for estimating divergence times integrating genomic and trait data

After one species splits into two, their genes and genomes will evolve independently. If the rate of evolution is constant over time, the genetic differences between species will accumulate at a fixed pace, proportional to the time of species separation. Thus molecules can serve as a clock, keeping time of species separation by the accumulated changes. If fossil records or geological events can be used to assign an absolute geological time to a species divergence event on the evolutionary tree, one can convert all calculated genetic distances into absolute geological times. This rationale for molecular clock dating has recently been extended to deal with variable evolutionary rate over time in the so-called relaxed-clock models. For fast-evolving viral DNA, the different sampling times of the viral sequences allows us to similarly calibrate the molecular and to obtain estimates of the absolute divergence times and evolutionary rates. In this project, we will develop statistical models of trait evolution, which will be used to analyse morphological trait data for both living and extinct species to generate fossil calibrations, which are crucially important to molecular dating analysis. Such models of trait evolution will also allow us to study the correlation between viral molecular sequence evolution and viral genotypes such as antigenic drift. We will apply the new methods to analyse real datasets to date major divergence events in the tree of life, such as the divergences of the human and the apes, the primates, the animals, and the flowering plants.

Molecular clock dating methods have been improved recently to accommodate the violation of the clock by the use of relaxed-clock models and to incorporate uncertainties in fossil calibrations through the use of soft bounds. Yet, representation of errors and uncertainties in the fossil record in a molecular dating analysis remains a challenging task. In this project, we will implement models of trait evolution to conduct Bayesian MCMC analysis of morphological traits in fossil and extent species. The resulting posterior for divergence times will be used as calibration densities for molecular clock dating. The new models and methods will be implemented in the MCMCtree program in the paml package, and will be applied to large datasets to date divergence events in the metazoan, the hominoid and primates, and the flowering plants. We will also analyse skull measurements of fossil and modern species within the hominoids, to generate posterior estimates of the hominoid divergence times, which will be used in a multispecies coalescent analysis of the hominoid genomic sequence data, to generate estimates of human-chimpanzee divergence time and of the mutation rate. Our mutation rate estimates will be vitally important to testing hypotheses concerning the origin and migration patterns of modern humans. We will use the same trait-evolution models to analyse viral phenotype (such as influenza virus epitopes) and its correlation with the evolutionary rate of the bird flu protein hemagglutinin.

We will implement the methods and algorithms to be developed in this project in the MCMCTREE program in the PAML software package, and distribute it at its web site, free of charge to academics. We will also develop a project-specific website including YouTube-hosted video manuals for the software.

We will attend local and international meetings to present our research results. Methodological advances will be disseminated in this way, as well as through teaching in the world-leading MSc Palaeobiology at Bristol, and the advanced workshop on Computational Molecular Evolution (funded by the Wellcome Trust and the EMBO) that is organized and co-instructed by Yang.