From trees to networks and back

Bastkowski, Sarah (2013) From trees to networks and back. Doctoral thesis, University of East Anglia.

[img]
Preview
PDF
Download (2MB) | Preview

Abstract

The evolutionary history of a set of species is commonly represented by a phylogenetic tree. Often, however, the data contain conflicting signals, which can be better represented by a more general structure, namely a phylogenetic network. Such networks allow the display of
several alternative evolutionary scenarios simultaneously but this can come at the price of complex visual representations. Using so-called circular split networks reduces this complexity, because this type of network can always be visualized in the plane without any crossing
edges. These circular split networks form the core of this thesis. We construct them, use them as a search space for minimum evolution trees and explore their properties.
More specifically, we present a new method, called SuperQ, to construct a circular split network summarising a collection of phylogenetic trees that have overlapping leaf sets. Then, we explore the set of phylogenetic trees associated with a �fixed circular split network, in particular using it as a search space for optimal trees. This set
represents just a tiny fraction of the space of all phylogenetic trees, but we still �find trees within it that compare quite favourably with those obtained by a leading heuristic, which uses tree edit operations for searching the whole tree space. In the last part, we advance our
understanding of the set of phylogenetic trees associated with a circular split network. Specifically, we investigate the size of the so-called circular tree neighbourhood for the three tree edit operations, tree bisection and reconnection (tbr), subtree prune and regraft (spr) and nearest neighbour interchange (nni).

Item Type: Thesis (Doctoral)
Faculty \ School: Faculty of Science > School of Computing Sciences
Depositing User: Brian Watkins
Date Deposited: 17 Jul 2014 13:21
Last Modified: 04 Jun 2015 00:38
URI: https://ueaeprints.uea.ac.uk/id/eprint/49692
DOI:

Actions (login required)

View Item View Item