New algorithms for the classification and visualisation of evolutionary and biomedical data

点击翻译按钮获取中文摘要

My research proposal includes four major parts: 1. Species evolution has long been modeled using only phylogenetic trees. In such a tree, each species has a unique most recent ancestor, whereas other interspecies relationships, such as those caused by horizontal gene transfers or hybridization, cannot be represented. First, I propose to develop two methods, based on the distance and probabilistic approaches respectively, for the prediction and visualization of possible horizontal gene transfer events. The complete and partial gene transfer models allowing for a number of important evolutionary constraints will be considered. 2. Second, I intend to develop a method for the analysis and representation of reticulate evolution by hybrid networks. I will design an algorithm and corresponding software for computing the smallest number of hybridization events required to explain conflicting gene trees and develop an algorithm to determine whether an arbitrary rooted network can be realized by contemporaneous reticulation events. 3. Third, I propose to continue the investigation of the problem of phylogenetic inference from sequence data containing missing bases. A new effective approach allowing one to estimate unknown nucleotides prior to computing the evolutionary distances between them will be designed and tested in simulations. 4. My fourth objective consists of designing and implementing methods for studying and correcting high-throughput screening (HTS) data. High-throughput screening is an effective technology for drug discovery. I plan to develop two methods for adjusting the hit selection procedure during HTS experiments. The proposed techniques will allow researchers to minimize the impact of systematic error affecting the selection of potential drug targets.