Moore, N.C.A. and Prosser, P. (2008) The ultrametric constraint and its application to phylogenetics. Journal of Artificial Intelligence Research, 32, pp. 901-938. (doi: 10.1613/jair.2580)
Full text not currently available from Enlighten.
Publisher's URL: http://jair.org/papers/paper2580.html
Abstract
A phylogenetic tree shows the evolutionary relationships among species. Internal nodes of the tree represent speciation events and leaf nodes correspond to species. A goal of phylogenetics is to combine such trees into larger trees, called supertrees, whilst respecting the relationships in the original trees. A rooted tree exhibits an ultrametric property; that is, for any three leaves of the tree it must be that one pair has a deeper most recent common ancestor than the other pairs, or that all three have the same most recent common ancestor. This inspires a constraint programming encoding for rooted trees. We present an efficient constraint that enforces the ultrametric property over a symmetric array of constrained integer variables, with the inevitable property that the lower bounds of any three variables are mutually supportive. We show that this allows an efficient constraint-based solution to the supertree construction problem. We demonstrate that the versatility of constraint programming can be exploited to allow solutions to variants of the supertree construction problem.
Item Type: | Articles |
---|---|
Status: | Published |
Refereed: | Yes |
Glasgow Author(s) Enlighten ID: | Prosser, Dr Patrick |
Authors: | Moore, N.C.A., and Prosser, P. |
College/School: | College of Science and Engineering > School of Computing Science |
Journal Name: | Journal of Artificial Intelligence Research |
Journal Abbr.: | J. artif. intell. res. |
Publisher: | AAAI Press |
ISSN: | 1076-9757 |
ISSN (Online): | 1943-5037 |
University Staff: Request a correction | Enlighten Editors: Update this record