Unifying Parsimonious Tree Reconciliation

TL;DR

This paper introduces a generalized framework for phylogenetic tree reconciliation that covers various coevolutionary systems, providing efficient algorithms for maximum parsimony reconciliation and an implementation tool.

Contribution

It presents a unified model for coevolutionary event reconciliation and develops both heuristic and exact algorithms with practical implementation.

Findings

Efficient O(n^2) heuristic algorithm for maximum parsimony reconciliation.

Exact branch-and-bound algorithm improves reconciliation accuracy.

Open-source Java tool available for practical use.

Abstract

Evolution is a process that is influenced by various environmental factors, e.g. the interactions between different species, genes, and biogeographical properties. Hence, it is interesting to study the combined evolutionary history of multiple species, their genes, and the environment they live in. A common approach to address this research problem is to describe each individual evolution as a phylogenetic tree and construct a tree reconciliation which is parsimonious with respect to a given event model. Unfortunately, most of the previous approaches are designed only either for host-parasite systems, for gene tree/species tree reconciliation, or biogeography. Hence, a method is desirable, which addresses the general problem of mapping phylogenetic trees and covering all varieties of coevolving systems, including e.g., predator-prey and symbiotic relationships. To overcome this gap, we introduce a generalized cophylogenetic event model considering the combinatorial complete set of local coevolutionary events. We give a dynamic programming based heuristic for solving the maximum parsimony reconciliation problem in time O(n^2), for two phylogenies each with at most n leaves. Furthermore, we present an exact branch-and-bound algorithm which uses the results from the dynamic programming heuristic for discarding partial reconciliations. The approach has been implemented as a Java application which is freely available from http://pacosy.informatik.uni-leipzig.de/coresym.