New algorithms for inferring multiple alternative consensus trees and supertrees

点击翻译按钮获取中文摘要

Each gene has its own evolutionary history which can substantially differ from the evolutionary histories of other genes. For example, some individual genes can be affected by specific horizontal gene transfer and recombination events. Thus, the evolutionary history of each gene should be represented by its own phylogenetic tree which may display different evolutionary patterns from the species tree (or Tree of Life) that accounts for the main patterns of vertical descent. Thus, the problem of inference of multiple consensus trees and supertrees becomes relevant. These alternative consensus trees and supertrees represent the most important reticulate evolutionary patterns characterizing a given set of gene phylogenies. They are often be much better resolved than a single consensus tree or a supertree inferred by traditional methods. Hence, using multiple consensus trees or supertrees we have the potential to extract much more plausible information from the given set of gene phylogenies. The goal of this computational biology research program is to establish the mathematical and statistical foundation for solving the challenging problem of classification of phylogenetic trees, and to create a new open-source platform allowing biologists use our new methods. We will define new optimization criteria for building multiple alternative consensus trees and supertrees using clustering algorithms and several important tree metrics. We propose the three following main axes for creating a computational framework for studying the evolutionary history of genes: (1) developing new effective algorithms to classify and build a plausible set of multiple alternative consensus trees and supertrees characterizing the data at hand, (2) exploring the patterns of evolution of Aminoacyl-tRNA synthetases, and (3) designing and implementing a new open-source software platform containing all the algorithms developed in the framework of this research program along with the real and simulated data. The long-term goal of this research program is to develop an innovative systematic and automated approach to reconstruct different scenarios of genetic evolution of different groups of species affected by reticulate evolutionary events. For example, the proposed research will contribute to the inference of multiple alternative subtrees of the Tree of Life that assembles the key information on the biodiversity and evolutionary relationships of all living organisms.