Estimating Divergence Times and Evolutionary Rates

点击翻译按钮获取中文摘要

There are almost as many divergences on the tree of life as there are extant species. Each of these divergences has a date (or, an interval presumed shorter in duration than the resultant lineages), and the fossil record is a priceless source of information about these dates. Although the fossil record is incomplete, the information that it provides can be supplemented with information extracted from DNA and protein sequence data. The basic idea is that molecular sequences should be similar if sequences are derived from closely related species and should be more diverged if the sequences come from more distantly related species. In this project, statistical procedures will be developed for combining fossil evidence and DNA or protein sequence data from extant organisms in order to estimate time since common ancestry. Special attention will be given to the fact that chronological rates of DNA and protein sequence evolution are not constant. Another goal of this project will be to infer the pattern by which these chronological rates of evolution change over time. Progress will be made by partitioning the change over time of chronological rates of molecular evolution into a portion that affects all genes in a lineage and a portion that affects each individual gene. Improved treatment of fossil data and handling of gene duplications will also be emphasized. The statistical procedures to be developed in this project will benefit a wide range of topics related to both biodiversity and molecular biology. The assignment of dates to branching events on the tree of life provides a framework for studying processes such as speciation and the dispersal of pathogens or other species. The ability to estimate chronological rates of molecular evolution and how these rates change over time also has diverse applications that include the study of protein function and functional innovation as well as the assessment of the importance to evolution of natural selection and genetic drift. Training for a graduate student in modern computational biology will be provided during the course of the research, and both the principal investigator, Dr. Jeffrey Thorne, and the graduate student will travel to Japan to work with the eminent statistician Dr. Hirohisa Kishino at the University of Tokyo. Source code developed as part of the project is made freely available, as has been the case with previous divergence-time software developed by Dr. Thorne.