Comp Bio: Collaborative Research: EMT: "Efficient Techniques for Reconstructing Horizontal Gene Transfer in Bacteria"

William Marsh Rice University, Houston TX

Investigators

Abstract

In eukaryotic organisms, evolution is mostly a process of vertical descent; in other words, genetic material is inherited from ancestors to descendants. In bacteria, on the other hand, genetic material may also be transferred "horizontally" among distantly related organisms. This process, known as horizontal gene transfer (HGT) and which is believed to be ubiquitous among various groups of bacteria, plays a major role in genome diversification as well as other bacteria-related issues, such as antibiotic resistance acquirement. In this project, we propose a set of methodologies to enable efficient and accurate phylogeny-based reconstruction of HGT in bacterial genome. In the phylogeny-based approach, a tree is built for each gene in the genome of a group of bacteria, and the gene tree disagreements are quantified and analyzed to detect HGT. Two major challenges face this approach: 1. Gene trees may disagree not only due to HGT, but also due to other processes, such as gene duplication and loss, lineage sorting, etc. Hence, it is imperative to develop a framework that determines the cause of gene tree disagreements before any estimates of HGT events are made. 2. Comparing and reconciling gene trees to obtain estimates of HGT events and rates are computationally very hard tasks. To ameliorate these two challenges, we propose to develop a stochastic framework that incorporates population genetics theories with evolutionary events that act among species to classify gene tree disagreements. Further, we will develop mathematical criteria and algorithmic techniques for comparing gene trees, quantifying their disagreements, and inferring the numbers and locations of HGT events. All the models and algorithms will be implemented in a software package, Sequoia, which will be made publicly available through open source mechanisms.